2023
- D. M. Karine Miras and A. E. Eiben, “Hu-bot: promoting the cooperation between humans and mobile robots.” 2023.
[BibTeX] [Abstract] [Download PDF]
This paper investigates human–robot collaboration in a novel setup: a human helps a mobile robot that can move and navigate freely in an environment. Specifically, the human helps by remotely taking over control during the learning of a task. The task is to find and collect several items in a walled arena, and Reinforcement Learning is used to seek a suitable controller. If the human observes undesired robot behavior, they can directly issue commands for the wheels through a game joystick. Experiments in a simulator showed that human assistance improved robot behavior efficacy by 30% and efficiency by 12%. The best policies were also tested in real life, using physical robots. Hardware experiments showed no significant difference concerning the simulations, providing empirical validation of our approach in practice.
@inproceedings{miras23hubot, title={Hu-bot: promoting the cooperation between humans and mobile robots}, author={Karine Miras, Decebal Mocanu & A. E. Eiben }, journal={Neural Computing and Applications }, year={2023}, url = {https://link.springer.com/article/10.1007/s00521-022-08061-z}, abstract = { This paper investigates human–robot collaboration in a novel setup: a human helps a mobile robot that can move and navigate freely in an environment. Specifically, the human helps by remotely taking over control during the learning of a task. The task is to find and collect several items in a walled arena, and Reinforcement Learning is used to seek a suitable controller. If the human observes undesired robot behavior, they can directly issue commands for the wheels through a game joystick. Experiments in a simulator showed that human assistance improved robot behavior efficacy by 30% and efficiency by 12%. The best policies were also tested in real life, using physical robots. Hardware experiments showed no significant difference concerning the simulations, providing empirical validation of our approach in practice.} } - T. Kim, M. Cochez, V. François-Lavet, M. Neerincx, and P. Vossen, “A Machine with Short-Term, Episodic, and Semantic Memory Systems.” 2023. doi:10.48550/ARXIV.2212.02098
[BibTeX] [Abstract] [Download PDF]
Inspired by the cognitive science theory of the explicit human memory systems, we have modeled an agent with short-term, episodic, and semantic memory systems, each of which is modeled with a knowledge graph. To evaluate this system and analyze the behavior of this agent, we designed and released our own reinforcement learning agent environment, “the Room”, where an agent has to learn how to encode, store, and retrieve memories to maximize its return by answering questions. We show that our deep Q-learning based agent successfully learns whether a short-term memory should be forgotten, or rather be stored in the episodic or semantic memory systems. Our experiments indicate that an agent with human-like memory systems can outperform an agent without this memory structure in the environment.
@InProceedings{https://doi.org/10.48550/arxiv.2212.02098, title = {A Machine with Short-Term, Episodic, and Semantic Memory Systems}, author = {Kim, Taewoon and Cochez, Michael and François-Lavet, Vincent and Neerincx, Mark and Vossen, Piek}, doi = {10.48550/ARXIV.2212.02098}, url = {https://arxiv.org/abs/2212.02098}, year = {2023}, month = {Feb.}, journal = {Proceedings of the AAAI Conference on Artificial Intelligence}, abstract = {Inspired by the cognitive science theory of the explicit human memory systems, we have modeled an agent with short-term, episodic, and semantic memory systems, each of which is modeled with a knowledge graph. To evaluate this system and analyze the behavior of this agent, we designed and released our own reinforcement learning agent environment, "the Room", where an agent has to learn how to encode, store, and retrieve memories to maximize its return by answering questions. We show that our deep Q-learning based agent successfully learns whether a short-term memory should be forgotten, or rather be stored in the episodic or semantic memory systems. Our experiments indicate that an agent with human-like memory systems can outperform an agent without this memory structure in the environment.}, }
2022
- M. Tsfasman, K. Fenech, M. Tarvirdians, A. Lorincz, C. Jonker, and C. Oertel, “Towards Creating a Conversational Memory for Long-Term Meeting Support: Predicting Memorable Moments in Multi-Party Conversations through Eye-Gaze,” in Proceedings of the 2022 International Conference on Multimodal Interaction, New York, NY, USA, 2022, p. 94–104. doi:10.1145/3536221.3556613
[BibTeX] [Abstract] [Download PDF]
When working in a group, it is essential to understand each other’s viewpoints to increase group cohesion and meeting productivity. This can be challenging in teams: participants might be left misunderstood and the discussion could be going around in circles. To tackle this problem, previous research on group interactions has addressed topics such as dominance detection, group engagement, and group creativity. Conversational memory, however, remains a widely unexplored area in the field of multimodal analysis of group interaction. The ability to track what each participant or a group as a whole find memorable from each meeting would allow a system or agent to continuously optimise its strategy to help a team meet its goals. In the present paper, we therefore investigate what participants take away from each meeting and how it is reflected in group dynamics.As a first step toward such a system, we recorded a multimodal longitudinal meeting corpus (MEMO), which comprises a first-party annotation of what participants remember from a discussion and why they remember it. We investigated whether participants of group interactions encode what they remember non-verbally and whether we can use such non-verbal multimodal features to predict what groups are likely to remember automatically. We devise a coding scheme to cluster participants’ memorisation reasons into higher-level constructs. We find that low-level multimodal cues, such as gaze and speaker activity, can predict conversational memorability. We also find that non-verbal signals can indicate when a memorable moment starts and ends. We could predict four levels of conversational memorability with an average accuracy of 44 \%. We also showed that reasons related to participants’ personal feelings and experiences are the most frequently mentioned grounds for remembering meeting segments.
@inproceedings{10.1145/3536221.3556613, author = {Tsfasman, Maria and Fenech, Kristian and Tarvirdians, Morita and Lorincz, Andras and Jonker, Catholijn and Oertel, Catharine}, title = {Towards Creating a Conversational Memory for Long-Term Meeting Support: Predicting Memorable Moments in Multi-Party Conversations through Eye-Gaze}, year = {2022}, isbn = {9781450393904}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3536221.3556613}, doi = {10.1145/3536221.3556613}, abstract = {When working in a group, it is essential to understand each other’s viewpoints to increase group cohesion and meeting productivity. This can be challenging in teams: participants might be left misunderstood and the discussion could be going around in circles. To tackle this problem, previous research on group interactions has addressed topics such as dominance detection, group engagement, and group creativity. Conversational memory, however, remains a widely unexplored area in the field of multimodal analysis of group interaction. The ability to track what each participant or a group as a whole find memorable from each meeting would allow a system or agent to continuously optimise its strategy to help a team meet its goals. In the present paper, we therefore investigate what participants take away from each meeting and how it is reflected in group dynamics.As a first step toward such a system, we recorded a multimodal longitudinal meeting corpus (MEMO), which comprises a first-party annotation of what participants remember from a discussion and why they remember it. We investigated whether participants of group interactions encode what they remember non-verbally and whether we can use such non-verbal multimodal features to predict what groups are likely to remember automatically. We devise a coding scheme to cluster participants’ memorisation reasons into higher-level constructs. We find that low-level multimodal cues, such as gaze and speaker activity, can predict conversational memorability. We also find that non-verbal signals can indicate when a memorable moment starts and ends. We could predict four levels of conversational memorability with an average accuracy of 44 \%. We also showed that reasons related to participants’ personal feelings and experiences are the most frequently mentioned grounds for remembering meeting segments.}, booktitle = {Proceedings of the 2022 International Conference on Multimodal Interaction}, pages = {94–104}, numpages = {11}, keywords = {multi-party interaction, social signals, conversational memory, multi-modal corpora}, location = {Bengaluru, India}, series = {ICMI '22} } - B. Dudzik, D. Küster, D. St-Onge, and F. Putze, “The 4th Workshop on Modeling Socio-Emotional and Cognitive Processes from Multimodal Data In-the-Wild (MSECP-Wild),” in Proceedings of the 2022 International Conference on Multimodal Interaction, New York, NY, USA, 2022, p. 803–804. doi:10.1145/3536221.3564029
[BibTeX] [Abstract] [Download PDF]
The ability to automatically infer relevant aspects of human users’ thoughts and feelings is crucial for technologies to adapt their behaviors in complex interactions intelligently (e.g., social robots or tutoring systems). Research on multimodal analysis has demonstrated the potential of technology to provide such estimates for a broad range of internal states and processes. However, constructing robust enough approaches for deployment in real-world applications remains an open problem. The MSECP-Wild workshop series serves as a multidisciplinary forum to present and discuss research addressing this challenge. This 4th iteration focuses on addressing varying contextual conditions (e.g., throughout an interaction or across different situations and environments) in intelligent systems as a crucial barrier for more valid real-world predictions and actions. Submissions to the workshop span efforts relevant to multimodal data collection and context-sensitive modeling. These works provide important impulses for discussions of the state-of-the-art and opportunities for future research on these subjects.
@inproceedings{10.1145/3536221.3564029, author = {Dudzik, Bernd and K\"{u}ster, Dennis and St-Onge, David and Putze, Felix},title = {The 4th Workshop on Modeling Socio-Emotional and Cognitive Processes from Multimodal Data In-the-Wild (MSECP-Wild)}, year = {2022}, isbn = {9781450393904}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://pure.tudelft.nl/admin/files/140622118/3536221.3564029.pdf}, doi = {10.1145/3536221.3564029}, abstract = {The ability to automatically infer relevant aspects of human users’ thoughts and feelings is crucial for technologies to adapt their behaviors in complex interactions intelligently (e.g., social robots or tutoring systems). Research on multimodal analysis has demonstrated the potential of technology to provide such estimates for a broad range of internal states and processes. However, constructing robust enough approaches for deployment in real-world applications remains an open problem. The MSECP-Wild workshop series serves as a multidisciplinary forum to present and discuss research addressing this challenge. This 4th iteration focuses on addressing varying contextual conditions (e.g., throughout an interaction or across different situations and environments) in intelligent systems as a crucial barrier for more valid real-world predictions and actions. Submissions to the workshop span efforts relevant to multimodal data collection and context-sensitive modeling. These works provide important impulses for discussions of the state-of-the-art and opportunities for future research on these subjects.}, booktitle = {Proceedings of the 2022 International Conference on Multimodal Interaction}, pages = {803–804}, numpages = {2}, keywords = {Context-awareness, User-Modeling, Multimodal Data, Social Signal Processing, Affective Computing, Ubiquitous Computing}, location = {Bengaluru, India}, series = {ICMI '22} } - B. Dudzik and H. Hung, “Exploring the Detection of Spontaneous Recollections during Video-Viewing In-the-Wild Using Facial Behavior Analysis,” in Proceedings of the 2022 International Conference on Multimodal Interaction, New York, NY, USA, 2022, p. 236–246. doi:10.1145/3536221.3556609
[BibTeX] [Abstract] [Download PDF]
Intelligent systems might benefit from automatically detecting when a stimulus has triggered a user’s recollection of personal memories, e.g., to identify that a piece of media content holds personal significance for them. While computational research has demonstrated the potential to identify related states based on facial behavior (e.g., mind-wandering), the automatic detection of spontaneous recollections specifically has not been investigated this far. Motivated by this, we present machine learning experiments exploring the feasibility of detecting whether a video clip has triggered personal memories in a viewer based on the analysis of their Head Rotation, Head Position, Eye Gaze, and Facial Expressions. Concretely, we introduce an approach for automatic detection and evaluate its potential for predictions using in-the-wild webcam recordings. Overall, our findings demonstrate the capacity for above chance detections in both settings, with substantially better performance for the video-independent variant. Beyond this, we investigate the role of person-specific recollection biases for predictions of our video-independent models and the importance of specific modalities of facial behavior. Finally, we discuss the implications of our findings for detecting recollections and user-modeling in adaptive systems.
@inproceedings{10.1145/3536221.3556609, author = {Dudzik, Bernd and Hung, Hayley}, title = {Exploring the Detection of Spontaneous Recollections during Video-Viewing In-the-Wild Using Facial Behavior Analysis}, year = {2022}, isbn = {9781450393904}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://dl.acm.org/doi/10.1145/3536221.3556609}, doi = {10.1145/3536221.3556609}, abstract = {Intelligent systems might benefit from automatically detecting when a stimulus has triggered a user’s recollection of personal memories, e.g., to identify that a piece of media content holds personal significance for them. While computational research has demonstrated the potential to identify related states based on facial behavior (e.g., mind-wandering), the automatic detection of spontaneous recollections specifically has not been investigated this far. Motivated by this, we present machine learning experiments exploring the feasibility of detecting whether a video clip has triggered personal memories in a viewer based on the analysis of their Head Rotation, Head Position, Eye Gaze, and Facial Expressions. Concretely, we introduce an approach for automatic detection and evaluate its potential for predictions using in-the-wild webcam recordings. Overall, our findings demonstrate the capacity for above chance detections in both settings, with substantially better performance for the video-independent variant. Beyond this, we investigate the role of person-specific recollection biases for predictions of our video-independent models and the importance of specific modalities of facial behavior. Finally, we discuss the implications of our findings for detecting recollections and user-modeling in adaptive systems.}, booktitle = {Proceedings of the 2022 International Conference on Multimodal Interaction}, pages = {236–246}, numpages = {11}, keywords = {Memories, Affective Computing, Recollection, Cognitive Processing, User-Modeling, Mind-Wandering, Facial Behavior Analysis}, location = {Bengaluru, India}, series = {ICMI '22} } - E. van Krieken, E. Acar, and F. van Harmelen, “Analyzing differentiable fuzzy logic operators,” Artificial Intelligence, vol. 302, p. 103602, 2022.
[BibTeX] [Abstract] [Download PDF]
The AI community is increasingly putting its attention towards combining symbolic and neural approaches, as it is often argued that the strengths and weaknesses of these approaches are complementary. One recent trend in the literature are weakly supervised learning techniques that employ operators from fuzzy logics. In particular, these use prior background knowledge described in such logics to help the training of a neural network from unlabeled and noisy data. By interpreting logical symbols using neural networks, this background knowledge can be added to regular loss functions, hence making reasoning a part of learning. We study, both formally and empirically, how a large collection of logical operators from the fuzzy logic literature behave in a differentiable learning setting. We find that many of these operators, including some of the most well-known, are highly unsuitable in this setting. A further finding concerns the treatment of implication in these fuzzy logics, and shows a strong imbalance between gradients driven by the antecedent and the consequent of the implication. Furthermore, we introduce a new family of fuzzy implications (called sigmoidal implications) to tackle this phenomenon. Finally, we empirically show that it is possible to use Differentiable Fuzzy Logics for semi-supervised learning, and compare how different operators behave in practice. We find that, to achieve the largest performance improvement over a supervised baseline, we have to resort to non-standard combinations of logical operators which perform well in learning, but no longer satisfy the usual logical laws.
@article{van2022analyzing, title = {Analyzing differentiable fuzzy logic operators}, author = {van Krieken, Emile and Acar, Erman and van Harmelen, Frank}, journal = {Artificial Intelligence}, volume = {302}, pages = {103602}, year = {2022}, publisher = {Elsevier}, url = "https://research.vu.nl/ws/portalfiles/portal/146020254/2002.06100v2.pdf", abstract = "The AI community is increasingly putting its attention towards combining symbolic and neural approaches, as it is often argued that the strengths and weaknesses of these approaches are complementary. One recent trend in the literature are weakly supervised learning techniques that employ operators from fuzzy logics. In particular, these use prior background knowledge described in such logics to help the training of a neural network from unlabeled and noisy data. By interpreting logical symbols using neural networks, this background knowledge can be added to regular loss functions, hence making reasoning a part of learning. We study, both formally and empirically, how a large collection of logical operators from the fuzzy logic literature behave in a differentiable learning setting. We find that many of these operators, including some of the most well-known, are highly unsuitable in this setting. A further finding concerns the treatment of implication in these fuzzy logics, and shows a strong imbalance between gradients driven by the antecedent and the consequent of the implication. Furthermore, we introduce a new family of fuzzy implications (called sigmoidal implications) to tackle this phenomenon. Finally, we empirically show that it is possible to use Differentiable Fuzzy Logics for semi-supervised learning, and compare how different operators behave in practice. We find that, to achieve the largest performance improvement over a supervised baseline, we have to resort to non-standard combinations of logical operators which perform well in learning, but no longer satisfy the usual logical laws." } - D. W. Romero, A. Kuzina, E. J. Bekkers, J. M. Tomczak, and M. Hoogendoorn, “CKConv: Continuous Kernel Convolution For Sequential Data,” International Conference on Learning Representations (ICLR), 2022, 2022.
[BibTeX] [Abstract] [Download PDF]
Conventional neural architectures for sequential data present important limitations. Recurrent networks suffer from exploding and vanishing gradients, small effective memory horizons, and must be trained sequentially. Convolutional networks are unable to handle sequences of unknown size and their memory horizon must be defined a priori. In this work, we show that all these problems can be solved by formulating convolutional kernels in CNNs as continuous functions. The resulting Continuous Kernel Convolution (CKConv) allows us to model arbitrarily long sequences in a parallel manner, within a single operation, and without relying on any form of recurrence. We show that Continuous Kernel Convolutional Networks (CKCNNs) obtain state-of-the-art results in multiple datasets, e.g., permuted MNIST, and, thanks to their continuous nature, are able to handle non-uniformly sampled datasets and irregularly-sampled data natively. CKCNNs match or perform better than neural ODEs designed for these purposes in a faster and simpler manner.
@article{DBLP:journals/corr/abs-2102-02611, author = {David W. Romero and Anna Kuzina and Erik J. Bekkers and Jakub M. Tomczak and Mark Hoogendoorn}, title = {CKConv: Continuous Kernel Convolution For Sequential Data}, journal = {International Conference on Learning Representations (ICLR), 2022}, year = {2022}, url = {https://openreview.net/pdf?id=8FhxBtXSl0}, abstract = {Conventional neural architectures for sequential data present important limitations. Recurrent networks suffer from exploding and vanishing gradients, small effective memory horizons, and must be trained sequentially. Convolutional networks are unable to handle sequences of unknown size and their memory horizon must be defined a priori. In this work, we show that all these problems can be solved by formulating convolutional kernels in CNNs as continuous functions. The resulting Continuous Kernel Convolution (CKConv) allows us to model arbitrarily long sequences in a parallel manner, within a single operation, and without relying on any form of recurrence. We show that Continuous Kernel Convolutional Networks (CKCNNs) obtain state-of-the-art results in multiple datasets, e.g., permuted MNIST, and, thanks to their continuous nature, are able to handle non-uniformly sampled datasets and irregularly-sampled data natively. CKCNNs match or perform better than neural ODEs designed for these purposes in a faster and simpler manner.} } - F. Sarvi, M. Heuss, M. Aliannejadi, S. Schelter, and M. de Rijke, “Understanding and Mitigating the Effect of Outliers in Fair Ranking,” in WSDM 2022: The Fifteenth International Conference on Web Search and Data Mining, 2022.
[BibTeX] [Abstract] [Download PDF]
Traditional ranking systems are expected to sort items in the order of their relevance and thereby maximize their utility. In fair ranking, utility is complemented with fairness as an optimization goal. Recent work on fair ranking focuses on developing algorithms to optimize for fairness, given position-based exposure. In contrast, we identify the potential of outliers in a ranking to influence exposure and thereby negatively impact fairness. An outlier in a list of items can alter the examination probabilities, which can lead to different distributions of attention, compared to position-based exposure. We formalize outlierness in a ranking, show that outliers are present in realistic datasets, and present the results of an eye-tracking study, showing that users scanning order and the exposure of items are influenced by the presence of outliers. We then introduce OMIT, a method for fair ranking in the presence of outliers. Given an outlier detection method, OMIT improves fair allocation of exposure by suppressing outliers in the top-k ranking. Using an academic search dataset, we show that outlierness optimization leads to a fairer policy that displays fewer outliers in the top-k, while maintaining a reasonable trade-off between fairness and utility.
@inproceedings{sarvi-2022-understanding, author = {Sarvi, Fatemeh and Heuss, Maria and Aliannejadi, Mohammad and Schelter, Sebastian and de Rijke, Maarten}, booktitle = {WSDM 2022: The Fifteenth International Conference on Web Search and Data Mining}, month = {February}, publisher = {ACM}, title = {Understanding and Mitigating the Effect of Outliers in Fair Ranking}, year = {2022}, url = {https://arxiv.org/abs/2112.11251}, abstract = "Traditional ranking systems are expected to sort items in the order of their relevance and thereby maximize their utility. In fair ranking, utility is complemented with fairness as an optimization goal. Recent work on fair ranking focuses on developing algorithms to optimize for fairness, given position-based exposure. In contrast, we identify the potential of outliers in a ranking to influence exposure and thereby negatively impact fairness. An outlier in a list of items can alter the examination probabilities, which can lead to different distributions of attention, compared to position-based exposure. We formalize outlierness in a ranking, show that outliers are present in realistic datasets, and present the results of an eye-tracking study, showing that users scanning order and the exposure of items are influenced by the presence of outliers. We then introduce OMIT, a method for fair ranking in the presence of outliers. Given an outlier detection method, OMIT improves fair allocation of exposure by suppressing outliers in the top-k ranking. Using an academic search dataset, we show that outlierness optimization leads to a fairer policy that displays fewer outliers in the top-k, while maintaining a reasonable trade-off between fairness and utility." } - E. Liscio, M. van der Meer, L. C. Siebert, C. M. Jonker, and P. K. Murukannaiah, “What values should an agent align with?,” Autonomous Agents and Multi-Agent Systems, vol. 36, iss. 23, p. 32, 2022.
[BibTeX] [Abstract] [Download PDF]
The pursuit of values drives human behavior and promotes cooperation. Existing research is focused on general values (e.g., Schwartz) that transcend contexts. However, context-specific values are necessary to (1) understand human decisions, and (2) engineer intelligent agents that can elicit and align with human values. We propose Axies, a hybrid (human and AI) methodology to identify context-specific values. Axies simplifies the abstract task of value identification as a guided value annotation process involving human annotators. Axies exploits the growing availability of value-laden text corpora and Natural Language Processing to assist the annotators in systematically identifying context-specific values. We evaluate Axies in a user study involving 80 human subjects. In our study, six annotators generate value lists for two timely and important contexts: Covid-19 measures and sustainable Energy. We employ two policy experts and 72 crowd workers to evaluate Axies value lists and compare them to a list of general (Schwartz) values. We find that Axies yields values that are (1) more context-specific than general values, (2) more suitable for value annotation than general values, and (3) independent of the people applying the methodology.
@article{Liscio2022, author = {Liscio, Enrico and van der Meer, Michiel and Siebert, Luciano C. and Jonker, Catholijn M. and Murukannaiah, Pradeep K.}, url = {https://link.springer.com/content/pdf/10.1007/s10458-022-09550-0}, journal = {Autonomous Agents and Multi-Agent Systems}, number = {23}, pages = {32}, publisher = {Springer US}, title = {{What values should an agent align with?}}, volume = {36}, year = {2022}, abstract = {The pursuit of values drives human behavior and promotes cooperation. Existing research is focused on general values (e.g., Schwartz) that transcend contexts. However, context-specific values are necessary to (1) understand human decisions, and (2) engineer intelligent agents that can elicit and align with human values. We propose Axies, a hybrid (human and AI) methodology to identify context-specific values. Axies simplifies the abstract task of value identification as a guided value annotation process involving human annotators. Axies exploits the growing availability of value-laden text corpora and Natural Language Processing to assist the annotators in systematically identifying context-specific values. We evaluate Axies in a user study involving 80 human subjects. In our study, six annotators generate value lists for two timely and important contexts: Covid-19 measures and sustainable Energy. We employ two policy experts and 72 crowd workers to evaluate Axies value lists and compare them to a list of general (Schwartz) values. We find that Axies yields values that are (1) more context-specific than general values, (2) more suitable for value annotation than general values, and (3) independent of the people applying the methodology.} } - G. Nadizar, E. Medvet, and K. Miras, “On the Schedule for Morphological Development of Evolved Modular Soft Robots,” in European Conference on Genetic Programming (Part of EvoStar), 2022, p. 146–161.
[BibTeX] [Abstract] [Download PDF]
Development is fundamental for living beings. As robots are often designed to mimic biological organisms, development is believed to be crucial for achieving successful results in robotic agents, as well. What is not clear, though, is the most appropriate scheduling for development. While in real life systems development happens mostly during the initial growth phase of organisms, it has not yet been investigated whether such assumption holds also for artificial creatures. In this paper, we employ a evolutionary approach to optimize the development – according to different representations – of Voxel-based Soft Robots (VSRs), a kind of modular robots. In our study, development consists in the addition of new voxels to the VSR, at fixed time instants, depending on the development schedule. We experiment with different schedules and show that, similarly to living organisms, artificial agents benefit from development occurring at early stages of life more than from development lasting for their entire life.
@inproceedings{nadizar2022schedule, title = {On the Schedule for Morphological Development of Evolved Modular Soft Robots}, author = {Nadizar, Giorgia and Medvet, Eric and Miras, Karine}, booktitle = {European Conference on Genetic Programming (Part of EvoStar)}, pages = {146--161}, year = {2022}, organization = {Springer}, url = {https://link.springer.com/chapter/10.1007/978-3-031-02056-8_10}, abstract = {Development is fundamental for living beings. As robots are often designed to mimic biological organisms, development is believed to be crucial for achieving successful results in robotic agents, as well. What is not clear, though, is the most appropriate scheduling for development. While in real life systems development happens mostly during the initial growth phase of organisms, it has not yet been investigated whether such assumption holds also for artificial creatures. In this paper, we employ a evolutionary approach to optimize the development - according to different representations - of Voxel-based Soft Robots (VSRs), a kind of modular robots. In our study, development consists in the addition of new voxels to the VSR, at fixed time instants, depending on the development schedule. We experiment with different schedules and show that, similarly to living organisms, artificial agents benefit from development occurring at early stages of life more than from development lasting for their entire life.} } - J. Kiseleva, Z. Li, M. Aliannejadi, S. Mohanty, M. ter Hoeve, M. Burtsev, A. Skrynnik, A. Zholus, A. Panov, K. Srinet, A. Szlam, Y. Sun, K. Hofmann Marc-Alexandre Côté, A. Awadallah, L. Abdrazakov, I. Churin, P. Manggala, K. Naszadi, M. van der Meer, and T. Kim, “Interactive Grounded Language Understanding in a Collaborative Environment: IGLU 2021,” , 2022. doi:10.48550/ARXIV.2205.02388
[BibTeX] [Abstract] [Download PDF]
Human intelligence has the remarkable ability to quickly adapt to new tasks and environments. Starting from a very young age, humans acquire new skills and learn how to solve new tasks either by imitating the behavior of others or by following provided natural language instructions. To facilitate research in this direction, we propose \emph{IGLU: Interactive Grounded Language Understanding in a Collaborative Environment}. The primary goal of the competition is to approach the problem of how to build interactive agents that learn to solve a task while provided with grounded natural language instructions in a collaborative environment. Understanding the complexity of the challenge, we split it into sub-tasks to make it feasible for participants.
@article{IGLU2022, author = {Kiseleva, Julia and Li, Ziming and Aliannejadi, Mohammad and Mohanty, Shrestha and ter Hoeve, Maartje and Burtsev, Mikhail and Skrynnik, Alexey and Zholus, Artem and Panov, Aleksandr and Srinet, Kavya and Szlam, Arthur and Sun, Yuxuan and Hofmann, Marc-Alexandre Côté, Katja and Awadallah, Ahmed and Abdrazakov, Linar and Churin, Igor and Manggala, Putra and Naszadi, Kata and van der Meer, Michiel and Kim, Taewoon}, keywords = {Computation and Language (cs.CL), Artificial Intelligence (cs.AI), FOS: Computer and information sciences, FOS: Computer and information sciences}, title = {Interactive Grounded Language Understanding in a Collaborative Environment: IGLU 2021}, publisher = {arXiv}, year = {2022}, abstract = "Human intelligence has the remarkable ability to quickly adapt to new tasks and environments. Starting from a very young age, humans acquire new skills and learn how to solve new tasks either by imitating the behavior of others or by following provided natural language instructions. To facilitate research in this direction, we propose \emph{IGLU: Interactive Grounded Language Understanding in a Collaborative Environment}. The primary goal of the competition is to approach the problem of how to build interactive agents that learn to solve a task while provided with grounded natural language instructions in a collaborative environment. Understanding the complexity of the challenge, we split it into sub-tasks to make it feasible for participants.", year = {2022}, doi = {10.48550/ARXIV.2205.02388}, url = {https://arxiv.org/abs/2205.02388}, copyright = {Creative Commons Attribution Share Alike 4.0 International} } - D. Grossi, “Social Choice Around the Block: On the Computational Social Choice of Blockchain,” in 21st International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2022, Auckland, New Zealand, May 9-13, 2022, 2022, p. 1788–1793.
[BibTeX] [Abstract] [Download PDF]
One of the most innovative aspects of blockchain technology con- sists in the introduction of an incentive layer to regulate the behav- ior of distributed protocols. The designer of a blockchain system faces therefore issues that are akin to those relevant for the design of economic mechanisms, and faces them in a computational setting. From this perspective the present paper argues for the importance of computational social choice in blockchain research. It identifies a few challenges at the interface of the two fields that illustrate the strong potential for cross-fertilization between them.
@inproceedings{DBLP:conf/atal/Grossi22, author = {Davide Grossi}, editor = {Piotr Faliszewski and Viviana Mascardi and Catherine Pelachaud and Matthew E. Taylor}, title = {Social Choice Around the Block: On the Computational Social Choice of Blockchain}, booktitle = {21st International Conference on Autonomous Agents and Multiagent Systems, {AAMAS} 2022, Auckland, New Zealand, May 9-13, 2022}, pages = {1788--1793}, publisher = {International Foundation for Autonomous Agents and Multiagent Systems {(IFAAMAS)}}, year = {2022}, url = {https://www.ifaamas.org/Proceedings/aamas2022/pdfs/p1788.pdf}, abstract = {One of the most innovative aspects of blockchain technology con- sists in the introduction of an incentive layer to regulate the behav- ior of distributed protocols. The designer of a blockchain system faces therefore issues that are akin to those relevant for the design of economic mechanisms, and faces them in a computational setting. From this perspective the present paper argues for the importance of computational social choice in blockchain research. It identifies a few challenges at the interface of the two fields that illustrate the strong potential for cross-fertilization between them.} } - M. G. Atigh, J. Schoep, E. Acar, N. van Noord, and P. Mettes, “Hyperbolic Image Segmentation,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 4453-4462.
[BibTeX] [Abstract] [Download PDF]
For image segmentation, the current standard is to perform pixel-level optimization and inference in Euclidean output embedding spaces through linear hyperplanes. In this work, we show that hyperbolic manifolds provide a valuable alternative for image segmentation and propose a tractable formulation of hierarchical pixel-level classification in hyperbolic space. Hyperbolic Image Segmentation opens up new possibilities and practical benefits for segmentation, such as uncertainty estimation and boundary information for free, zero-label generalization, and increased performance in low-dimensional output embeddings.
@InProceedings{Atigh_2022_CVPR, author = {Atigh, Mina Ghadimi and Schoep, Julian and Acar, Erman and van Noord, Nanne and Mettes, Pascal}, title = {Hyperbolic Image Segmentation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2022}, pages = {4453-4462}, url = {https://openaccess.thecvf.com/content/CVPR2022/papers/Atigh_Hyperbolic_Image_Segmentation_CVPR_2022_paper.pdf}, abstract = {For image segmentation, the current standard is to perform pixel-level optimization and inference in Euclidean output embedding spaces through linear hyperplanes. In this work, we show that hyperbolic manifolds provide a valuable alternative for image segmentation and propose a tractable formulation of hierarchical pixel-level classification in hyperbolic space. Hyperbolic Image Segmentation opens up new possibilities and practical benefits for segmentation, such as uncertainty estimation and boundary information for free, zero-label generalization, and increased performance in low-dimensional output embeddings.} } - R. Verma and E. Nalisnick, “Calibrated Learning to Defer with One-vs-All Classifiers,” in ICML 2022 Workshop on Human-Machine Collaboration and Teaming, 2022.
[BibTeX] [Abstract] [Download PDF]
The learning to defer (L2D) framework has the potential to make AI systems safer. For a given input, the system can defer the decision to a human if the human is more likely than the model to take the correct action. We study the calibration of L2D systems, investigating if the probabilities they output are sound. We find that Mozannar & Sontag’s (2020) multiclass framework is not calibrated with respect to expert correctness. Moreover, it is not even guaranteed to produce valid probabilities due to its parameterization being degenerate for this purpose. We propose an L2D system based on one-vs-all classifiers that is able to produce calibrated probabilities of expert correctness. Furthermore, our loss function is also a consistent surrogate for multiclass L2D, like Mozannar & Sontag’s (2020). Our experiments verify that not only is our system calibrated, but this benefit comes at no cost to accuracy. Our model’s accuracy is always comparable (and often superior) to Mozannar & Sontag’s (2020) model’s in tasks ranging from hate speech detection to galaxy classification to diagnosis of skin lesions.
@inproceedings{Verma-Nalisnick-ICML:2022, title = {Calibrated Learning to Defer with One-vs-All Classifiers}, author = {Rajeev Verma and Eric Nalisnick}, year = {2022}, booktitle = {ICML 2022 Workshop on Human-Machine Collaboration and Teaming}, abstract = {The learning to defer (L2D) framework has the potential to make AI systems safer. For a given input, the system can defer the decision to a human if the human is more likely than the model to take the correct action. We study the calibration of L2D systems, investigating if the probabilities they output are sound. We find that Mozannar & Sontag’s (2020) multiclass framework is not calibrated with respect to expert correctness. Moreover, it is not even guaranteed to produce valid probabilities due to its parameterization being degenerate for this purpose. We propose an L2D system based on one-vs-all classifiers that is able to produce calibrated probabilities of expert correctness. Furthermore, our loss function is also a consistent surrogate for multiclass L2D, like Mozannar & Sontag's (2020). Our experiments verify that not only is our system calibrated, but this benefit comes at no cost to accuracy. Our model's accuracy is always comparable (and often superior) to Mozannar & Sontag's (2020) model's in tasks ranging from hate speech detection to galaxy classification to diagnosis of skin lesions.}, url = {https://icml.cc/Conferences/2022/ScheduleMultitrack?event=18123} } - P. Manggala, H. H. Hoos, and E. Nalisnick, “Bayesian Weak Supervision via an Optimal Transport Approach,” in ICML 2022 Workshop on Human-Machine Collaboration and Teaming, 2022.
[BibTeX] [Abstract] [Download PDF]
Large-scale machine learning is often impeded by a lack of labeled training data. To address this problem, the paradigm of weak supervision aims to collect and then aggregate multiple noisy labels. We propose a Bayesian probabilistic model that employs a tractable Sinkhorn-based optimal transport formulation to derive a ground-truth label. The translation between true and weak labels is cast as a transport problem with an inferred cost structure. Our approach achieves strong performance on the WRENCH weak supervision benchmark. Moreover, the posterior distribution over cost matrices allows for exploratory analysis of the weak sources.
@inproceedings{manggala2022optimaltransportweaksupervision, title = {Bayesian Weak Supervision via an Optimal Transport Approach}, author = {Manggala, Putra and Hoos, Holger H. and Nalisnick, Eric}, year = {2022}, booktitle = {ICML 2022 Workshop on Human-Machine Collaboration and Teaming}, abstract = {Large-scale machine learning is often impeded by a lack of labeled training data. To address this problem, the paradigm of weak supervision aims to collect and then aggregate multiple noisy labels. We propose a Bayesian probabilistic model that employs a tractable Sinkhorn-based optimal transport formulation to derive a ground-truth label. The translation between true and weak labels is cast as a transport problem with an inferred cost structure. Our approach achieves strong performance on the WRENCH weak supervision benchmark. Moreover, the posterior distribution over cost matrices allows for exploratory analysis of the weak sources.}, url = {https://openreview.net/forum?id=YJkf-6tTFiY} } - R. Dobbe, “System Safety and Artificial Intelligence,” in Oxford Handbook of AI Governance, Oxford: , 2022, vol. To Appear.
[BibTeX] [Abstract] [Download PDF]
This chapter formulates seven lessons for preventing harm in artificial intelligence (AI) systems based on insights from the field of system safety for software-based automation in safety-critical domains. New applications of AI across societal domains and public organizations and infrastructures come with new hazards, which lead to new forms of harm, both grave and pernicious. The text addresses the lack of consensus for diagnosing and eliminating new AI system hazards. For decades, the field of system safety has dealt with accidents and harm in safety-critical systems governed by varying degrees of software-based automation and decision-making. This field embraces the core assumption of systems and control that AI systems cannot be safeguarded by technical design choices on the model or algorithm alone, instead requiring an end-to-end hazard analysis and design frame that includes the context of use, impacted stakeholders and the formal and informal institutional environment in which the system operates. Safety and other values are then inherently socio-technical and emergent system properties that require design and control measures to instantiate these across the technical, social and institutional components of a system. This chapter honors system safety pioneer Nancy Leveson, by situating her core lessons for today’s AI system safety challenges. For every lesson, concrete tools are offered for rethinking and reorganizing the safety management of AI systems, both in design and governance. This history tells us that effective AI safety management requires transdisciplinary approaches and a shared language that allows involvement of all levels of society.
@incollection{dobbe_system_2022, address = {Oxford}, title = {System {Safety} and {Artificial} {Intelligence}}, volume = {To Appear}, isbn = {978-0-19-757932-9}, url = {https://arxiv.org/abs/2202.09292}, abstract = {This chapter formulates seven lessons for preventing harm in artificial intelligence (AI) systems based on insights from the field of system safety for software-based automation in safety-critical domains. New applications of AI across societal domains and public organizations and infrastructures come with new hazards, which lead to new forms of harm, both grave and pernicious. The text addresses the lack of consensus for diagnosing and eliminating new AI system hazards. For decades, the field of system safety has dealt with accidents and harm in safety-critical systems governed by varying degrees of software-based automation and decision-making. This field embraces the core assumption of systems and control that AI systems cannot be safeguarded by technical design choices on the model or algorithm alone, instead requiring an end-to-end hazard analysis and design frame that includes the context of use, impacted stakeholders and the formal and informal institutional environment in which the system operates. Safety and other values are then inherently socio-technical and emergent system properties that require design and control measures to instantiate these across the technical, social and institutional components of a system. This chapter honors system safety pioneer Nancy Leveson, by situating her core lessons for today's AI system safety challenges. For every lesson, concrete tools are offered for rethinking and reorganizing the safety management of AI systems, both in design and governance. This history tells us that effective AI safety management requires transdisciplinary approaches and a shared language that allows involvement of all levels of society.}, booktitle = {Oxford {Handbook} of {AI} {Governance}}, author = {Dobbe, Roel}, year = {2022} } - A. Sauter, E. Acar, and V. François-Lavet, A Meta-Reinforcement Learning Algorithm for Causal DiscoveryarXiv, 2022. doi:10.48550/ARXIV.2207.08457
[BibTeX] [Abstract] [Download PDF]
Causal discovery is a major task with the utmost importance for machine learning since causal structures can enable models to go beyond pure correlation-based inference and significantly boost their performance. However, finding causal structures from data poses a significant challenge both in computational effort and accuracy, let alone its impossibility without interventions in general. In this paper, we develop a meta-reinforcement learning algorithm that performs causal discovery by learning to perform interventions such that it can construct an explicit causal graph. Apart from being useful for possible downstream applications, the estimated causal graph also provides an explanation for the data-generating process. In this article, we show that our algorithm estimates a good graph compared to the SOTA approaches, even in environments whose underlying causal structure is previously unseen. Further, we make an ablation study that shows how learning interventions contribute to the overall performance of our approach. We conclude that interventions indeed help boost the performance, efficiently yielding an accurate estimate of the causal structure of a possibly unseen environment.
@misc{Sauter22MetaRL, doi = {10.48550/ARXIV.2207.08457}, url = {https://arxiv.org/abs/2207.08457}, author = {Sauter, Andreas and Acar, Erman and François-Lavet, Vincent}, title = {A Meta-Reinforcement Learning Algorithm for Causal Discovery}, publisher = {arXiv}, year = {2022}, copyright = {Creative Commons Attribution Share Alike 4.0 International}, abstract = { Causal discovery is a major task with the utmost importance for machine learning since causal structures can enable models to go beyond pure correlation-based inference and significantly boost their performance. However, finding causal structures from data poses a significant challenge both in computational effort and accuracy, let alone its impossibility without interventions in general. In this paper, we develop a meta-reinforcement learning algorithm that performs causal discovery by learning to perform interventions such that it can construct an explicit causal graph. Apart from being useful for possible downstream applications, the estimated causal graph also provides an explanation for the data-generating process. In this article, we show that our algorithm estimates a good graph compared to the SOTA approaches, even in environments whose underlying causal structure is previously unseen. Further, we make an ablation study that shows how learning interventions contribute to the overall performance of our approach. We conclude that interventions indeed help boost the performance, efficiently yielding an accurate estimate of the causal structure of a possibly unseen environment.} } - M. Heuss, F. Sarvi, and M. de Rijke, “Fairness of Exposure in Light of Incomplete Exposure Estimation,” in Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, New York, NY, USA, 2022, pp. 759-769.
[BibTeX] [Abstract] [Download PDF]
Fairness of exposure is a commonly used notion of fairness for ranking systems. It is based on the idea that all items or item groups should get exposure proportional to the merit of the item or the collective merit of the items in the group. Often, stochastic ranking policies are used to ensure fairness of exposure. Previous work unrealistically assumes that we can reliably estimate the expected exposure for all items in each ranking produced by the stochastic policy. In this work, we discuss how to approach fairness of exposure in cases where the policy contains rankings of which, due to inter-item dependencies, we cannot reliably estimate the exposure distribution. In such cases, we cannot determine whether the policy can be considered fair. Our contributions in this paper are twofold. First, we define a method called method for finding stochastic policies that avoid showing rankings with unknown exposure distribution to the user without having to compromise user utility or item fairness. Second, we extend the study of fairness of exposure to the top-k setting and also assess method in this setting. We find that method can significantly reduce the number of rankings with unknown exposure distribution without a drop in user utility or fairness compared to existing fair ranking methods, both for full-length and top-k rankings. This is an important first step in developing fair ranking methods for cases where we have incomplete knowledge about the user’s behaviour.
@inproceedings{heuss-2022-fairness, author = {Heuss, Maria and Sarvi, Fatemeh and de Rijke, Maarten}, title = {Fairness of Exposure in Light of Incomplete Exposure Estimation}, year = {2022}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://irlab.science.uva.nl/wp-content/papercite-data/pdf/heuss-2022-fairness.pdf}, booktitle = {Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval}, pages = {759-769}, location = {Madrid, Spain}, abstract = {Fairness of exposure is a commonly used notion of fairness for ranking systems. It is based on the idea that all items or item groups should get exposure proportional to the merit of the item or the collective merit of the items in the group. Often, stochastic ranking policies are used to ensure fairness of exposure. Previous work unrealistically assumes that we can reliably estimate the expected exposure for all items in each ranking produced by the stochastic policy. In this work, we discuss how to approach fairness of exposure in cases where the policy contains rankings of which, due to inter-item dependencies, we cannot reliably estimate the exposure distribution. In such cases, we cannot determine whether the policy can be considered fair. Our contributions in this paper are twofold. First, we define a method called method for finding stochastic policies that avoid showing rankings with unknown exposure distribution to the user without having to compromise user utility or item fairness. Second, we extend the study of fairness of exposure to the top-k setting and also assess method in this setting. We find that method can significantly reduce the number of rankings with unknown exposure distribution without a drop in user utility or fairness compared to existing fair ranking methods, both for full-length and top-k rankings. This is an important first step in developing fair ranking methods for cases where we have incomplete knowledge about the user's behaviour.} } - L. Ho, V. de Boer, B. M. van Riemsdijk, S. Schlobach, and M. Tielman, “Argumentation for Knowledge Base Inconsistencies in Hybrid Intelligence Scenarios,” , online, 2022.
[BibTeX] [Abstract] [Download PDF]
Hybrid Intelligence (HI) is the combination of human and machine intelligence, expanding human intellect instead of replacing it. In HI scenarios, inconsistencies in knowledge bases (KBs) can occur for a variety of reasons. These include shifting preferences, user’s motivation and or external conditions (for example, available resources and environment can vary over time). Argumentation is a potential method to address such inconsistencies as it provides a mechanism for reasoning with conflicting information, with natural explanations that are understandable to humans. In this paper, we investigate the capabilities of Argumentation in representing and reasoning about knowledge of both human and artificial agents in the presence of inconsistency. Moreover, we show how Argumentation enables Explainability for addressing problems in Decision-Making and Justification of an opinion. In order to investigate the applicability of Argumentation in HI scenarios, we demonstrate a mapping of two specific HI scenarios to Argumentation problems. We analyse to what extent of Argumentation is applicable by clarifying the practical inconsistency types of the HI scenarios that Argumentation can address. These include inconsistencies related to recommendations and decision making. We then model particularly the presentation of conflicting information for each scenario based on the form of argument representation.
@inproceedings{HHAI2022, address = {online}, author = {Ho, Loan and de Boer, Victor and van Riemsdijk, M. Birna and Schlobach, Stefan and Tielman, Myrthe}, publisher = {The 1st International Workshop on Knowledge Representation for Hybrid Intelligence (KR4HI)}, title = {Argumentation for Knowledge Base Inconsistencies in Hybrid Intelligence Scenarios}, year = {2022}, url = "https://drive.google.com/file/d/1QD95kbvzej0mXCxMQDkHOctklqCjMWO6/view", abstract = "Hybrid Intelligence (HI) is the combination of human and machine intelligence, expanding human intellect instead of replacing it. In HI scenarios, inconsistencies in knowledge bases (KBs) can occur for a variety of reasons. These include shifting preferences, user’s motivation and or external conditions (for example, available resources and environment can vary over time). Argumentation is a potential method to address such inconsistencies as it provides a mechanism for reasoning with conflicting information, with natural explanations that are understandable to humans. In this paper, we investigate the capabilities of Argumentation in representing and reasoning about knowledge of both human and artificial agents in the presence of inconsistency. Moreover, we show how Argumentation enables Explainability for addressing problems in Decision-Making and Justification of an opinion. In order to investigate the applicability of Argumentation in HI scenarios, we demonstrate a mapping of two specific HI scenarios to Argumentation problems. We analyse to what extent of Argumentation is applicable by clarifying the practical inconsistency types of the HI scenarios that Argumentation can address. These include inconsistencies related to recommendations and decision making. We then model particularly the presentation of conflicting information for each scenario based on the form of argument representation.", keywords = {Hybrid Intelligence, Knowledge Representation and Reasoning, Argumentation, Explainability, Inconsistency, Preferences} } - L. Ho, S. Arch-int, E. Acar, S. Schlobach, and N. Arch-int, “An argumentative approach for handling inconsistency in prioritized Datalog± ontologies,” AI Commun., vol. 35, iss. 3, pp. 243-267, 2022. doi:10.3233/AIC-220087
[BibTeX] [Abstract] [Download PDF]
Prioritized Datalog± is a well-studied formalism for modelling ontological knowledge and data, and has a success story in many applications in the (Semantic) Web and in other domains. Since the information content on the Web is both inherently context-dependent and frequently updated, the occurrence of a logical inconsistency is often inevitable. This phenomenon has led the research community to develop various types of inconsistency-tolerant semantics over the last few decades. Although the study of query answering under inconsistency-tolerant semantics is well-understood, the problem of explaining query answering under such semantics took considerably less attention, especially in the scenario where the facts are prioritized. In this paper, we aim to fill this gap. More specifically, we use Dung”s abstract argumentation framework to address the problem of explaining inconsistency-tolerant query answering in Datalog± KB where facts are prioritized, or preordered. We clarify the relationship between preferred repair semantics and various notions of extensions for argumentation frameworks. The strength of such argumentation-based approach is the explainability; users can more easily understand why different points of views are conflicting and why the query answer is entailed (or not) under different semantics. To this end we introduce the formal notion of a dialogical explanation, and show how it can be used to both explain showing why query results hold and not hold according to the known semantics in inconsistent Datalog± knowledge bases.
@article{10.3233/AIC-220087, author = {Ho, Loan and Arch-int, Somjit and Acar, Erman and Schlobach, Stefan and Arch-int, Ngamnij}, title = {An argumentative approach for handling inconsistency in prioritized Datalog± ontologies}, year = {2022}, journal = {AI Commun.}, month = {jan}, pages = {243-267}, numpages = {25}, keywords = {Argumentation, Datalog±, inconsistency, preferences, prioritized knowledge bases, explanation}, issue_date = {2022}, publisher = {IOS Press}, address = {NLD}, volume = {35}, number = {3}, issn = {0921-7126}, url = {https://doi.org/10.3233/AIC-220087}, doi = {10.3233/AIC-220087}, abstract = {Prioritized Datalog± is a well-studied formalism for modelling ontological knowledge and data, and has a success story in many applications in the (Semantic) Web and in other domains. Since the information content on the Web is both inherently context-dependent and frequently updated, the occurrence of a logical inconsistency is often inevitable. This phenomenon has led the research community to develop various types of inconsistency-tolerant semantics over the last few decades. Although the study of query answering under inconsistency-tolerant semantics is well-understood, the problem of explaining query answering under such semantics took considerably less attention, especially in the scenario where the facts are prioritized. In this paper, we aim to fill this gap. More specifically, we use Dung''s abstract argumentation framework to address the problem of explaining inconsistency-tolerant query answering in Datalog± KB where facts are prioritized, or preordered. We clarify the relationship between preferred repair semantics and various notions of extensions for argumentation frameworks. The strength of such argumentation-based approach is the explainability; users can more easily understand why different points of views are conflicting and why the query answer is entailed (or not) under different semantics. To this end we introduce the formal notion of a dialogical explanation, and show how it can be used to both explain showing why query results hold and not hold according to the known semantics in inconsistent Datalog± knowledge bases.} } - L. Ho, V. de Boer, B. M. van Riemsdijk, S. Schlobach, and M. Tielman, “Knowledge Representation Formalisms for Hybrid Intelligence,” , online, 2022.
[BibTeX] [Abstract] [Download PDF]
Hybrid Intelligence (HI) is the combination of human and machine intelligence, expanding human intellect instead of replacing it. Information in HI scenarios is often inconsistent, e.g. due to shifting preferences, user’s motivation or conflicts arising from merged data. As it provides an intuitive mechanism for reasoning with conflicting information, with natural explanations that are understandable to humans, our hypothesis is that Dung’s Abstract Argumentation (AA) is a suitable formalism for such hybrid scenarios. This paper investigates the capabilities of Argumentation in representing and reasoning in the presence of inconsistency, and its potential for intuitive explainability to link between artificial and human actors. To this end, we conduct a survey among a number of research projects of the Hybrid Intelligence Centre1 . Within these projects we analyse the applicability of argumentation with respect to various inconsistency types stemming, for instance, from commonsense reasoning, decision making, and negotiation. The results show that 14 out of the 21 projects have to deal with inconsistent information. In half of those scenarios, the knowledge models come with natural preference relations over the information. We show that Argumentation is a suitable framework to model the specific knowledge in 10 out of 14 projects, thus indicating the potential of Abstract Argumentation for transparently dealing with inconsistencies in Hybrid Intelligence systems.
@inproceedings{ArgXAI2022, address = {online}, author = {Ho, Loan and de Boer, Victor and van Riemsdijk, M. Birna and Schlobach, Stefan and Tielman, Myrthe}, publisher = {1st International Workshop on Argumentation for eXplainable AI (ArgXAI) co-located with 9th International Conference on Computational Models of Argument (COMMA 2022)}, title = {Knowledge Representation Formalisms for Hybrid Intelligence}, volume = {Vol-3209}, year = {2022}, url = "http://ceur-ws.org/Vol-3209/7787.pdf", month = {september}, abstract = "Hybrid Intelligence (HI) is the combination of human and machine intelligence, expanding human intellect instead of replacing it. Information in HI scenarios is often inconsistent, e.g. due to shifting preferences, user's motivation or conflicts arising from merged data. As it provides an intuitive mechanism for reasoning with conflicting information, with natural explanations that are understandable to humans, our hypothesis is that Dung's Abstract Argumentation (AA) is a suitable formalism for such hybrid scenarios. This paper investigates the capabilities of Argumentation in representing and reasoning in the presence of inconsistency, and its potential for intuitive explainability to link between artificial and human actors. To this end, we conduct a survey among a number of research projects of the Hybrid Intelligence Centre1 . Within these projects we analyse the applicability of argumentation with respect to various inconsistency types stemming, for instance, from commonsense reasoning, decision making, and negotiation. The results show that 14 out of the 21 projects have to deal with inconsistent information. In half of those scenarios, the knowledge models come with natural preference relations over the information. We show that Argumentation is a suitable framework to model the specific knowledge in 10 out of 14 projects, thus indicating the potential of Abstract Argumentation for transparently dealing with inconsistencies in Hybrid Intelligence systems.", keywords = {Hybrid Intelligence, Argumentation, Explainability, Inconsistency, Preferences} } - S. Renooij, “Relevance for Robust Bayesian Network MAP-Explanations,” in Proceedings of The 11th International Conference on Probabilistic Graphical Models, 2022, p. 13–24.
[BibTeX] [Abstract] [Download PDF]
In the context of explainable AI, the concept of MAP-independence was recently introduced as a means for conveying the (ir)relevance of intermediate nodes for MAP computations in Bayesian networks. In this paper, we further study the concept of MAP-independence, discuss methods for finding sets of relevant nodes, and suggest ways to use these in providing users with an explanation concerning the robustness of the MAP result.
@InProceedings{pmlr-v186-renooij22a, title = {Relevance for Robust Bayesian Network MAP-Explanations}, author = {Renooij, Silja}, booktitle = {Proceedings of The 11th International Conference on Probabilistic Graphical Models}, pages = {13--24}, year = {2022}, editor = {Salmerán, Antonio and Rumí, Rafael}, volume = {186}, series = {Proceedings of Machine Learning Research}, month = {05--07 Oct}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v186/renooij22a/renooij22a.pdf}, url = {https://proceedings.mlr.press/v186/renooij22a.html}, abstract = {In the context of explainable AI, the concept of MAP-independence was recently introduced as a means for conveying the (ir)relevance of intermediate nodes for MAP computations in Bayesian networks. In this paper, we further study the concept of MAP-independence, discuss methods for finding sets of relevant nodes, and suggest ways to use these in providing users with an explanation concerning the robustness of the MAP result.} } - H. Prakken and R. Ratsma, “A top-level model of case-based argumentation for explanation: Formalisation and experiments,” Argument and Computation, vol. 13, pp. 159-194, 2022.
[BibTeX] [Abstract] [Download PDF]
This paper proposes a formal top-level model of explaining the outputs of machine-learning-based decision-making applications and evaluates it experimentally with three data sets. The model draws on AI & law research on argumentation with cases, which models how lawyers draw analogies to past cases and discuss their relevant similarities and differences in terms of relevant factors and dimensions in the problem domain. A case-based approach is natural since the input data of machine-learning applications can be seen as cases. While the approach is motivated by legal decision making, it also applies to other kinds of decision making, such as commercial decisions about loan applications or employee hiring, as long as the outcome is binary and the input conforms to this factor- or dimension format. The model is top-level in that it can be extended with more refined accounts of similarities and differences between cases. It is shown to overcome several limitations of similar argumentation-based explanation models, which only have binary features and do not represent the tendency of features towards particular outcomes. The results of the experimental evaluation studies indicate that the model may be feasible in practice, but that further development and experimentation is needed to confirm its usefulness as an explanation model. Main challenges here are selecting from a large number of possible explanations, reducing the number of features in the explanations and adding more meaningful information to them. It also remains to be investigated how suitable our approach is for explaining non-linear models.
@ARTICLE{p+r22, AUTHOR = "H. Prakken and R. Ratsma", TITLE = "A top-level model of case-based argumentation for explanation: Formalisation and experiments", JOURNAL = "Argument and Computation", YEAR = "2022", VOLUME = "13", PAGES = "159-194", Abstract = "This paper proposes a formal top-level model of explaining the outputs of machine-learning-based decision-making applications and evaluates it experimentally with three data sets. The model draws on AI & law research on argumentation with cases, which models how lawyers draw analogies to past cases and discuss their relevant similarities and differences in terms of relevant factors and dimensions in the problem domain. A case-based approach is natural since the input data of machine-learning applications can be seen as cases. While the approach is motivated by legal decision making, it also applies to other kinds of decision making, such as commercial decisions about loan applications or employee hiring, as long as the outcome is binary and the input conforms to this factor- or dimension format. The model is top-level in that it can be extended with more refined accounts of similarities and differences between cases. It is shown to overcome several limitations of similar argumentation-based explanation models, which only have binary features and do not represent the tendency of features towards particular outcomes. The results of the experimental evaluation studies indicate that the model may be feasible in practice, but that further development and experimentation is needed to confirm its usefulness as an explanation model. Main challenges here are selecting from a large number of possible explanations, reducing the number of features in the explanations and adding more meaningful information to them. It also remains to be investigated how suitable our approach is for explaining non-linear models.", URL = "https://content.iospress.com/articles/argument-and-computation/aac210009", } - M. M. Çelikok, F. A. Oliehoek, and S. Kaski, “Best-Response Bayesian Reinforcement Learning with Bayes-Adaptive POMDPs for Centaurs,” in Proceedings of the 21st International Conference on Autonomous Agents and Multiagent Systems, Richland, SC, 2022, pp. 235-243.
[BibTeX] [Abstract] [Download PDF]
Centaurs are half-human, half-AI decision-makers where the AI’s goal is to complement the human. To do so, the AI must be able to recognize the goals and constraints of the human and have the means to help them. We present a novel formulation of the interaction between the human and the AI as a sequential game where the agents are modelled using Bayesian best-response models. We show that in this case the AI’s problem of helping bounded-rational humans make better decisions reduces to a Bayes-adaptive POMDP. In our simulated experiments, we consider an instantiation of our framework for humans who are subjectively optimistic about the AI’s future behaviour. Our results show that when equipped with a model of the human, the AI can infer the human’s bounds and nudge them towards better decisions. We discuss ways in which the machine can learn to improve upon its own limitations as well with the help of the human. We identify a novel trade-off for centaurs in partially observable tasks: for the AI’s actions to be acceptable to the human, the machine must make sure their beliefs are sufficiently aligned, but aligning beliefs might be costly. We present a preliminary theoretical analysis of this trade-off and its dependence on task structure.
@inproceedings{10.5555/3535850.3535878, author = {Çelikok, Mustafa Mert and Oliehoek, Frans A. and Kaski, Samuel}, title = {Best-Response Bayesian Reinforcement Learning with Bayes-Adaptive POMDPs for Centaurs}, year = {2022}, isbn = {9781450392136}, publisher = {International Foundation for Autonomous Agents and Multiagent Systems}, address = {Richland, SC}, booktitle = {Proceedings of the 21st International Conference on Autonomous Agents and Multiagent Systems}, pages = {235-243}, numpages = {9}, keywords = {Bayesian reinforcement learning, multiagent learning, computational rationality, hybrid intelligence}, location = {Virtual Event, New Zealand}, series = {AAMAS '22}, url = {"https://dl.acm.org/doi/abs/10.5555/3535850.3535878"}, abstract = {Centaurs are half-human, half-AI decision-makers where the AI's goal is to complement the human. To do so, the AI must be able to recognize the goals and constraints of the human and have the means to help them. We present a novel formulation of the interaction between the human and the AI as a sequential game where the agents are modelled using Bayesian best-response models. We show that in this case the AI's problem of helping bounded-rational humans make better decisions reduces to a Bayes-adaptive POMDP. In our simulated experiments, we consider an instantiation of our framework for humans who are subjectively optimistic about the AI's future behaviour. Our results show that when equipped with a model of the human, the AI can infer the human's bounds and nudge them towards better decisions. We discuss ways in which the machine can learn to improve upon its own limitations as well with the help of the human. We identify a novel trade-off for centaurs in partially observable tasks: for the AI's actions to be acceptable to the human, the machine must make sure their beliefs are sufficiently aligned, but aligning beliefs might be costly. We present a preliminary theoretical analysis of this trade-off and its dependence on task structure.} } - H. Prakken, “Formalising an aspect of argument strength: degrees of attackability,” in Computational Models of Argument. Proceedings of COMMA 2022, F. T. et al., Ed., Amsterdam etc: IOS Press, 2022, pp. 296-307. doi:10.3233/FAIA220161
[BibTeX] [Abstract] [Download PDF]
This paper formally studies a notion of dialectical argument strength in terms of the number of ways in which an argument can be successfully attacked in expansions of an abstract argumentation framework. The proposed model is abstract but its design is motivated by the wish to avoid overly limiting assumptions that may not hold in particular dialogue contexts or in particular structured accounts of argumentation. It is shown that most principles for gradual argument acceptability proposed in the literature fail to hold for the proposed notion of dialectical strength, which clarifies their rational foundations and highlights the importance of distinguishing between logical, dialectical and rhetorical argument strength.
@INCOLLECTION{hp22gradual, AUTHOR = "H. Prakken", TITLE = "Formalising an aspect of argument strength: degrees of attackability", BOOKTITLE = "Computational Models of Argument. Proceedings of {COMMA} 2022", EDITOR = "Francesca Toni et al.", PUBLISHER = "IOS Press", ADDRESS = "Amsterdam etc", PAGES = "296-307", YEAR = "2022", abstract = "This paper formally studies a notion of dialectical argument strength in terms of the number of ways in which an argument can be successfully attacked in expansions of an abstract argumentation framework. The proposed model is abstract but its design is motivated by the wish to avoid overly limiting assumptions that may not hold in particular dialogue contexts or in particular structured accounts of argumentation. It is shown that most principles for gradual argument acceptability proposed in the literature fail to hold for the proposed notion of dialectical strength, which clarifies their rational foundations and highlights the importance of distinguishing between logical, dialectical and rhetorical argument strength.", DOI = "10.3233/FAIA220161", url = "https://ebooks.iospress.nl/doi/10.3233/FAIA220161" } - W. van Woerkom, D. Grossi, H. Prakken, and B. Verheij, “Justification in Case-Based Reasoning,” in Proceedings of the First International Workshop on Argumentation for eXplainable AI, 2022, pp. 1-13.
[BibTeX] [Abstract] [Download PDF]
The explanation and justification of decisions is an important subject in contemporary data-driven automated methods. Case-based argumentation has been proposed as the formal background for the explanation of data-driven automated decision making. In particular, a method was developed in recent work based on the theory of precedential constraint which reasons from a case base, given by the training data of the machine learning system, to produce a justification for the outcome of a focus case. An important role is played in this method by the notions of citability and compensation, and in the present work we develop these in more detail. Special attention is paid to the notion of compensation; we formally specify the notion and identify several of its desirable properties. These considerations reveal a refined formal perspective on the explanation method as an extension of the theory of precedential constraint with a formal notion of justification.
@inproceedings{vanwoerkom2022, author = "van Woerkom, Wijnand and Grossi, Davide and Prakken, Henry and Verheij, Bart", editor = "Čyras, Kristijonas and Kampik, Timotheus and Cocarascu, Oana and Rago, Antonio", title = "Justification in Case-Based Reasoning", booktitle = "Proceedings of the First International Workshop on Argumentation for eXplainable AI", year = "2022", pages = {1-13}, publisher = "CEUR Workshop Proceedings", Abstract = "The explanation and justification of decisions is an important subject in contemporary data-driven automated methods. Case-based argumentation has been proposed as the formal background for the explanation of data-driven automated decision making. In particular, a method was developed in recent work based on the theory of precedential constraint which reasons from a case base, given by the training data of the machine learning system, to produce a justification for the outcome of a focus case. An important role is played in this method by the notions of citability and compensation, and in the present work we develop these in more detail. Special attention is paid to the notion of compensation; we formally specify the notion and identify several of its desirable properties. These considerations reveal a refined formal perspective on the explanation method as an extension of the theory of precedential constraint with a formal notion of justification.", URL = "http://ceur-ws.org/Vol-3209/", } - N. Hopner, I. Tiddi, and H. van Hoof, “Leveraging Class Abstraction for Commonsense Reinforcement Learning via Residual Policy Gradient Methods,” in Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22, 2022, p. 3050–3056. doi:10.24963/ijcai.2022/423
[BibTeX] [Abstract] [Download PDF]
Enabling reinforcement learning (RL) agents to leverage a knowledge base while learning from experience promises to advance RL in knowledge intensive domains. However, it has proven difficult to leverage knowledge that is not manually tailored to the environment. We propose to use the subclass relationships present in open-source knowledge graphs to abstract away from specific objects. We develop a residual policy gradient method that is able to integrate knowledge across different abstraction levels in the class hierarchy. Our method results in improved sample efficiency and generalisation to unseen objects in commonsense games, but we also investigate failure modes, such as excessive noise in the extracted class knowledge or environments with little class structure.
@inproceedings{ijcai2022p423, title = {Leveraging Class Abstraction for Commonsense Reinforcement Learning via Residual Policy Gradient Methods}, author = {Hopner, Niklas and Tiddi, Ilaria and van Hoof, Herke}, booktitle = {Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, {IJCAI-22}}, publisher = {International Joint Conferences on Artificial Intelligence Organization}, editor = {Lud De Raedt}, pages = {3050--3056}, year = {2022}, month = {7}, note = {Main Track}, doi = {10.24963/ijcai.2022/423}, url = {https://doi.org/10.24963/ijcai.2022/423}, abstract = {Enabling reinforcement learning (RL) agents to leverage a knowledge base while learning from experience promises to advance RL in knowledge intensive domains. However, it has proven difficult to leverage knowledge that is not manually tailored to the environment. We propose to use the subclass relationships present in open-source knowledge graphs to abstract away from specific objects. We develop a residual policy gradient method that is able to integrate knowledge across different abstraction levels in the class hierarchy. Our method results in improved sample efficiency and generalisation to unseen objects in commonsense games, but we also investigate failure modes, such as excessive noise in the extracted class knowledge or environments with little class structure.} } - S. Baez Santamaria, P. Vossen, and T. Baier, “Evaluating Agent Interactions Through Episodic Knowledge Graphs,” in Proceedings of the 1st Workshop on Customized Chat Grounding Persona and Knowledge, Gyeongju, Republic of Korea, 2022, p. 15–28.
[BibTeX] [Abstract] [Download PDF]
We present a new method based on episodic Knowledge Graphs (eKGs) for evaluating (multimodal) conversational agents in open domains. This graph is generated by interpreting raw signals during conversation and is able to capture the accumulation of knowledge over time. We apply structural and semantic analysis of the resulting graphs and translate the properties into qualitative measures. We compare these measures with existing automatic and manual evaluation metrics commonly used for conversational agents. Our results show that our Knowledge-Graph-based evaluation provides more qualitative insights into interaction and the agent{‘}s behavior.
@inproceedings{baez-santamaria-etal-2022-evaluating, title = "Evaluating Agent Interactions Through Episodic Knowledge Graphs", author = "Baez Santamaria, Selene and Vossen, Piek and Baier, Thomas", booktitle = "Proceedings of the 1st Workshop on Customized Chat Grounding Persona and Knowledge", month = "oct", year = "2022", address = "Gyeongju, Republic of Korea", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.ccgpk-1.3", pages = "15--28", abstract = "We present a new method based on episodic Knowledge Graphs (eKGs) for evaluating (multimodal) conversational agents in open domains. This graph is generated by interpreting raw signals during conversation and is able to capture the accumulation of knowledge over time. We apply structural and semantic analysis of the resulting graphs and translate the properties into qualitative measures. We compare these measures with existing automatic and manual evaluation metrics commonly used for conversational agents. Our results show that our Knowledge-Graph-based evaluation provides more qualitative insights into interaction and the agent{'}s behavior." } - M. van der Meer, M. Reuver, U. Khurana, L. Krause, and S. Santamaría, Will It Blend? Mixing Training Paradigms & Prompting for Argument Quality PredictionarXiv, 2022. doi:10.48550/ARXIV.2209.08966
[BibTeX] [Abstract] [Download PDF]
This paper describes our winning contribution to the Shared Task of the 9th Workshop on Argument Mining (2022). Our approach uses Large Language Models for the task of Argument Quality Prediction. We perform prompt engineering using GPT-3, and also investigate the training paradigms multi-task learning, contrastive learning, and intermediate-task training. We find that a mixed prediction setup outperforms single models. Prompting GPT-3 works best for predicting argument validity, and argument novelty is best estimated by a model trained using all three training paradigms.
@misc{https://doi.org/10.48550/arxiv.2209.08966, doi = {10.48550/ARXIV.2209.08966}, url = {https://arxiv.org/abs/2209.08966}, author = {van der Meer, Michiel and Reuver, Myrthe and Khurana, Urja and Krause, Lea and Santamaría, Selene Báez}, title = {Will It Blend? Mixing Training Paradigms & Prompting for Argument Quality Prediction}, publisher = {arXiv}, year = {2022}, abstract = {This paper describes our winning contribution to the Shared Task of the 9th Workshop on Argument Mining (2022). Our approach uses Large Language Models for the task of Argument Quality Prediction. We perform prompt engineering using GPT-3, and also investigate the training paradigms multi-task learning, contrastive learning, and intermediate-task training. We find that a mixed prediction setup outperforms single models. Prompting GPT-3 works best for predicting argument validity, and argument novelty is best estimated by a model trained using all three training paradigms.}, copyright = {Creative Commons Attribution Non Commercial Share Alike 4.0 International} } - K. Miras and A. Eiben, “How the History of Changing Environments Affects Traits of Evolvable Robot Populations,” Artificial life, vol. 28, iss. 2, p. 224–239, 2022.
[BibTeX] [Abstract] [Download PDF]
The environment is one of the key factors in the emergence of intelligent creatures, but it has received little attention within the Evolutionary Robotics literature. This article investigates the effects of changing environments on morphological and behavioral traits of evolvable robots. In particular, we extend a previous study by evolving robot populations under diverse changing-environment setups, varying the magnitude, frequency, duration, and dynamics of the changes. The results show that long-lasting effects of early generations occur not only when transitioning from easy to hard conditions, but also when going from hard to easy conditions. Furthermore, we demonstrate how the impact of environmental scaffolding is dependent on the nature of the environmental changes involved.
@article{miras2022history, title = {How the History of Changing Environments Affects Traits of Evolvable Robot Populations}, author = {Miras, Karine and Eiben, AE}, journal = {Artificial life}, volume = {28}, url = {https://research.vu.nl/en/publications/how-the-history-of-changing-environments-affects-traits-of-evolva}, number = {2}, pages = {224--239}, year = {2022}, publisher = {MIT Press One Broadway, 12th Floor, Cambridge, Massachusetts 02142, USA~…}, abstract = {The environment is one of the key factors in the emergence of intelligent creatures, but it has received little attention within the Evolutionary Robotics literature. This article investigates the effects of changing environments on morphological and behavioral traits of evolvable robots. In particular, we extend a previous study by evolving robot populations under diverse changing-environment setups, varying the magnitude, frequency, duration, and dynamics of the changes. The results show that long-lasting effects of early generations occur not only when transitioning from easy to hard conditions, but also when going from hard to easy conditions. Furthermore, we demonstrate how the impact of environmental scaffolding is dependent on the nature of the environmental changes involved.} } - J. Schoemaker and K. Miras, “The benefits of credit assignment in noisy video game environments,” in Artificial Life Conference Proceedings 34, 2022, p. 6.
[BibTeX] [Abstract] [Download PDF]
Both Evolutionary Algorithms (EAs) and Reinforcement Learning Algorithms (RLAs) have proven successful in policy optimisation tasks, however, there is scarce literature comparing their strengths and weaknesses. This makes it difficult to determine which group of algorithms is best suited for a task. This paper presents a comparison of two EAs and two RLAs in solving EvoMan – a video game playing benchmark. We test the algorithms both with and without noise introduction in the initialisation of multiple video game environments. We demonstrate that EAs reach a similar performance to RLAs in the static environments, but when noise is introduced the performance of EAs drops drastically while the performance of RLAs is much less affected.
@inproceedings{schoemaker2022benefits, title = {The benefits of credit assignment in noisy video game environments}, author = {Schoemaker, Jacob and Miras, Karine}, booktitle = {Artificial Life Conference Proceedings 34}, volume = {2022}, number = {1}, pages = {6}, url = {https://direct.mit.edu/isal/proceedings-pdf/isal/34/6/2035319/isal_a_00483.pdf}, year = {2022}, organization = {MIT Press One Rogers Street, Cambridge, MA 02142-1209, USA journals-info~…}, abstract = {Both Evolutionary Algorithms (EAs) and Reinforcement Learning Algorithms (RLAs) have proven successful in policy optimisation tasks, however, there is scarce literature comparing their strengths and weaknesses. This makes it difficult to determine which group of algorithms is best suited for a task. This paper presents a comparison of two EAs and two RLAs in solving EvoMan - a video game playing benchmark. We test the algorithms both with and without noise introduction in the initialisation of multiple video game environments. We demonstrate that EAs reach a similar performance to RLAs in the static environments, but when noise is introduced the performance of EAs drops drastically while the performance of RLAs is much less affected.} } - W. van Woerkom, D. Grossi, H. Prakken, and B. Verheij, “Landmarks in Case-Based Reasoning: From Theory to Data,” in HHAI2022: Augmenting Human Intellect, 2022, pp. 212-224.
[BibTeX] [Abstract] [Download PDF]
Widespread application of uninterpretable machine learning systems for sensitive purposes has spurred research into elucidating the decision making process of these systems. These efforts have their background in many different disciplines, one of which is the field of AI & law. In particular, recent works have observed that machine learning training data can be interpreted as legal cases. Under this interpretation the formalism developed to study case law, called the theory of precedential constraint, can be used to analyze the way in which machine learning systems draw on training data – or should draw on them – to make decisions. These works predominantly stay on the theoretical level, hence in the present work the formalism is evaluated on a real world dataset. Through this analysis we identify a significant new concept which we call landmark cases, and use it to characterize the types of datasets that are more or less suitable to be described by the theory.
@InProceedings{woerkom2022landmarks, author = {van Woerkom, Wijnand and Grossi, Davide and Prakken, Henry and Verheij, Bart}, title = {Landmarks in Case-Based Reasoning: From Theory to Data}, booktitle = {HHAI2022: Augmenting Human Intellect}, year = {2022}, editor = {Schlobach, Stefan and Pérez-Ortiz, María and Tielman, Myrthe}, volume = {354}, series = {Frontiers in Artificial Intelligence and Applications}, pages = {212-224}, publisher = {IOS Press}, Abstract = {Widespread application of uninterpretable machine learning systems for sensitive purposes has spurred research into elucidating the decision making process of these systems. These efforts have their background in many different disciplines, one of which is the field of AI & law. In particular, recent works have observed that machine learning training data can be interpreted as legal cases. Under this interpretation the formalism developed to study case law, called the theory of precedential constraint, can be used to analyze the way in which machine learning systems draw on training data – or should draw on them – to make decisions. These works predominantly stay on the theoretical level, hence in the present work the formalism is evaluated on a real world dataset. Through this analysis we identify a significant new concept which we call landmark cases, and use it to characterize the types of datasets that are more or less suitable to be described by the theory.}, URL = {https://ebooks.iospress.nl/volumearticle/60868} } - N. Kökciyan and P. Yolum, “Taking Situation-Based Privacy Decisions: Privacy Assistants Working with Humans,” in Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence (IJCAI), 2022, p. 703–709.
[BibTeX] [Abstract] [Download PDF]
Privacy on the Web is typically managed by giving consent to individual Websites for various aspects of data usage. This paradigm requires too much human effort and thus is impractical for Internet of Things (IoT) applications where humans interact with many new devices on a daily basis. Ideally, software privacy assistants can help by making privacy decisions in different situations on behalf of the users. To realize this, we propose an agent-based model for a privacy assistant. The model identifies the contexts that a situation implies and computes the trustworthiness of these contexts. Contrary to traditional trust models that capture trust in an entity by observing large number of interactions, our proposed model can assess the trustworthiness even if the user has not interacted with the particular device before. Moreover, our model can decide which situations are inherently ambiguous and thus can request the human to make the decision. We evaluate various aspects of the model using a real-life data set and report adjustments that are needed to serve different types of users well.
@inproceedings{pas-ijcai-2022, title = {Taking Situation-Based Privacy Decisions: Privacy Assistants Working with Humans}, author = {Kökciyan, Nadin and Yolum, Pinar}, booktitle = {Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence (IJCAI)}, pages = {703--709}, year = {2022}, month = {7}, abstract = {Privacy on the Web is typically managed by giving consent to individual Websites for various aspects of data usage. This paradigm requires too much human effort and thus is impractical for Internet of Things (IoT) applications where humans interact with many new devices on a daily basis. Ideally, software privacy assistants can help by making privacy decisions in different situations on behalf of the users. To realize this, we propose an agent-based model for a privacy assistant. The model identifies the contexts that a situation implies and computes the trustworthiness of these contexts. Contrary to traditional trust models that capture trust in an entity by observing large number of interactions, our proposed model can assess the trustworthiness even if the user has not interacted with the particular device before. Moreover, our model can decide which situations are inherently ambiguous and thus can request the human to make the decision. We evaluate various aspects of the model using a real-life data set and report adjustments that are needed to serve different types of users well.}, url = {https://www.ijcai.org/proceedings/2022/0099.pdf} } - O. Ulusoy and P. Yolum, “PANOLA: A Personal Assistant for Supporting Users in Preserving Privacy,” ACM Transactions on Internet Technology, vol. 22, iss. 1, 2022.
[BibTeX] [Abstract] [Download PDF]
Privacy is the right of individuals to keep personal information to themselves. When individuals use online systems, they should be given the right to decide what information they would like to share and what to keep private. When a piece of information pertains only to a single individual, preserving privacy is possible by providing the right access options to the user. However, when a piece of information pertains to multiple individuals, such as a picture of a group of friends or a collaboratively edited document, deciding how to share this information and with whom is challenging. The problem becomes more difficult when the individuals who are affected by the information have different, possibly conflicting privacy constraints. Resolving this problem requires a mechanism that takes into account the relevant individuals’ concerns to decide on the privacy configuration of information. Because these decisions need to be made frequently (i.e., per each piece of shared content), the mechanism should be automated. This article presents a personal assistant to help end-users with managing the privacy of their content. When some content that belongs to multiple users is about to be shared, the personal assistants of the users employ an auction-based privacy mechanism to regulate the privacy of the content. To do so, each personal assistant learns the preferences of its user over time and produces bids accordingly. Our proposed personal assistant is capable of assisting users with different personas and thus ensures that people benefit from it as they need it. Our evaluations over multiagent simulations with online social network content show that our proposed personal assistant enables privacy-respecting content sharing.
@article{panola-2022, author = {Ulusoy, Onuralp and Yolum, Pinar}, title = {{PANOLA}: A Personal Assistant for Supporting Users in Preserving Privacy}, year = {2022}, issue_date = {February 2022}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, volume = {22}, number = {1}, journal = {ACM Transactions on Internet Technology}, month = {February}, articleno = {27}, numpages = {32}, url = {https://doi.org/10.1145/3471187}, abstract = {Privacy is the right of individuals to keep personal information to themselves. When individuals use online systems, they should be given the right to decide what information they would like to share and what to keep private. When a piece of information pertains only to a single individual, preserving privacy is possible by providing the right access options to the user. However, when a piece of information pertains to multiple individuals, such as a picture of a group of friends or a collaboratively edited document, deciding how to share this information and with whom is challenging. The problem becomes more difficult when the individuals who are affected by the information have different, possibly conflicting privacy constraints. Resolving this problem requires a mechanism that takes into account the relevant individuals’ concerns to decide on the privacy configuration of information. Because these decisions need to be made frequently (i.e., per each piece of shared content), the mechanism should be automated. This article presents a personal assistant to help end-users with managing the privacy of their content. When some content that belongs to multiple users is about to be shared, the personal assistants of the users employ an auction-based privacy mechanism to regulate the privacy of the content. To do so, each personal assistant learns the preferences of its user over time and produces bids accordingly. Our proposed personal assistant is capable of assisting users with different personas and thus ensures that people benefit from it as they need it. Our evaluations over multiagent simulations with online social network content show that our proposed personal assistant enables privacy-respecting content sharing.} } - Gönül. Ayci, M. Şensoy, A. Özgür, and P. Yolum, “Uncertainty-Aware Personal Assistant for Making Personalized Privacy Decisions,” ACM Transactions on Internet Technology, 2022. doi:10.1145/3561820
[BibTeX] [Abstract] [Download PDF]
Many software systems, such as online social networks enable users to share information about themselves. While the action of sharing is simple, it requires an elaborate thought process on privacy: what to share, with whom to share, and for what purposes. Thinking about these for each piece of content to be shared is tedious. Recent approaches to tackle this problem build personal assistants that can help users by learning what is private over time and recommending privacy labels such as private or public to individual content that a user considers sharing. However, privacy is inherently ambiguous and highly personal. Existing approaches to recommend privacy decisions do not address these aspects of privacy sufficiently. Ideally, a personal assistant should be able to adjust its recommendation based on a given user, considering that user’s privacy understanding. Moreover, the personal assistant should be able to assess when its recommendation would be uncertain and let the user make the decision on her own. Accordingly, this paper proposes a personal assistant that uses evidential deep learning to classify content based on its privacy label. An important characteristic of the personal assistant is that it can model its uncertainty in its decisions explicitly, determine that it does not know the answer, and delegate from making a recommendation when its uncertainty is high. By factoring in user’s own understanding of privacy, such as risk factors or own labels, the personal assistant can personalize its recommendations per user. We evaluate our proposed personal assistant using a well-known data set. Our results show that our personal assistant can accurately identify uncertain cases, personalize them to its user’s needs, and thus helps users preserve their privacy well.
@article{pure-2022, author = {Ayci, Gönül and Şensoy, Murat and Özgür, Arzucan and Yolum, Pinar}, title = {Uncertainty-Aware Personal Assistant for Making Personalized Privacy Decisions}, year = {2022}, journal = {ACM Transactions on Internet Technology}, publisher = {Association for Computing Machinery}, url = {https://doi.org/10.1145/3561820}, doi = {10.1145/3561820}, abstract = {Many software systems, such as online social networks enable users to share information about themselves. While the action of sharing is simple, it requires an elaborate thought process on privacy: what to share, with whom to share, and for what purposes. Thinking about these for each piece of content to be shared is tedious. Recent approaches to tackle this problem build personal assistants that can help users by learning what is private over time and recommending privacy labels such as private or public to individual content that a user considers sharing. However, privacy is inherently ambiguous and highly personal. Existing approaches to recommend privacy decisions do not address these aspects of privacy sufficiently. Ideally, a personal assistant should be able to adjust its recommendation based on a given user, considering that user’s privacy understanding. Moreover, the personal assistant should be able to assess when its recommendation would be uncertain and let the user make the decision on her own. Accordingly, this paper proposes a personal assistant that uses evidential deep learning to classify content based on its privacy label. An important characteristic of the personal assistant is that it can model its uncertainty in its decisions explicitly, determine that it does not know the answer, and delegate from making a recommendation when its uncertainty is high. By factoring in user’s own understanding of privacy, such as risk factors or own labels, the personal assistant can personalize its recommendations per user. We evaluate our proposed personal assistant using a well-known data set. Our results show that our personal assistant can accurately identify uncertain cases, personalize them to its user’s needs, and thus helps users preserve their privacy well.}, note = {In press} } - E. Erdogan, F. Dignum, R. Verbrugge, and P. Yolum, “Abstracting Minds: Computational Theory of Mind for Human-Agent Collaboration,” in HHAI2022: Augmenting Human Intellect, IOS Press, 2022, p. 199–211.
[BibTeX] [Abstract] [Download PDF]
Theory of mind refers to the human ability to reason about mental content of other people such as beliefs, desires, and goals. In everyday life, people rely on their theory of mind to understand, explain, and predict the behaviour of others. Having a theory of mind is especially useful when people collaborate, since individuals can then reason on what the other individual knows as well as what reasoning they might do. Realization of hybrid intelligence, where an agent collaborates with a human, will require the agent to be able to do similar reasoning through computational theory of mind. Accordingly, this paper provides a mechanism for computational theory of mind based on abstractions of single beliefs into higher-level concepts. These concepts can correspond to social norms, roles, as well as values. Their use in decision making serves as a heuristic to choose among interactions, thus facilitating collaboration on decisions. Using examples from the medical domain, we demonstrate how having such a theory of mind enables an agent to interact with humans efficiently and can increase the quality of the decisions humans make.
@incollection{erdogan2022abstracting, title = {Abstracting Minds: Computational Theory of Mind for Human-Agent Collaboration}, author = {Erdogan, Emre and Dignum, Frank and Verbrugge, Rineke and Yolum, Pinar}, booktitle = {HHAI2022: Augmenting Human Intellect}, pages = {199--211}, year = {2022}, publisher = {IOS Press}, url = {http://dx.doi.org/10.3233/FAIA220199}, abstract = {Theory of mind refers to the human ability to reason about mental content of other people such as beliefs, desires, and goals. In everyday life, people rely on their theory of mind to understand, explain, and predict the behaviour of others. Having a theory of mind is especially useful when people collaborate, since individuals can then reason on what the other individual knows as well as what reasoning they might do. Realization of hybrid intelligence, where an agent collaborates with a human, will require the agent to be able to do similar reasoning through computational theory of mind. Accordingly, this paper provides a mechanism for computational theory of mind based on abstractions of single beliefs into higher-level concepts. These concepts can correspond to social norms, roles, as well as values. Their use in decision making serves as a heuristic to choose among interactions, thus facilitating collaboration on decisions. Using examples from the medical domain, we demonstrate how having such a theory of mind enables an agent to interact with humans efficiently and can increase the quality of the decisions humans make.} } - E. Erdogan, F. Dignum, R. Verbrugge, and P. Yolum, “Computational Theory of Mind for Human-Agent Coordination,” in Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XV, 2022, p. 92–108.
[BibTeX] [Abstract] [Download PDF]
In everyday life, people often depend on their theory of mind, i.e., their ability to reason about unobservable mental content of others to understand, explain, and predict their behaviour. Many agent-based models have been designed to develop computational theory of mind and analyze its effectiveness in various tasks and settings. However, most existing models are not generic (e.g., only applied in a given setting), not feasible (e.g., require too much information to be processed), or not human-inspired (e.g., do not capture the behavioral heuristics of humans). This hinders their applicability in many settings. Accordingly, we propose a new computational theory of mind, which captures the human decision heuristics of reasoning by abstracting individual beliefs about others. We specifically study computational affinity and show how it can be used in tandem with theory of mind reasoning when designing agent models for human-agent negotiation. We perform two-agent simulations to analyze the role of affinity in getting to agreements when there is a bound on the time to be spent for negotiating. Our results suggest that modeling affinity can ease the negotiation process by decreasing the number of rounds needed for an agreement as well as yield a higher benefit for agents with theory of mind reasoning.
@InProceedings{erdogan+2022, author = "Erdogan, Emre and Dignum, Frank and Verbrugge, Rineke and Yolum, Pinar", editor = "Ajmeri, Nirav and Morris Martin, Andreasa and Savarimuthu, Bastin Tony Roy", title = "Computational Theory of Mind for Human-Agent Coordination", booktitle = "Coordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XV", pages = "92--108", year = "2022", publisher = "Springer International Publishing", url = "http://dx.doi.org/10.1007/978-3-031-20845-4_6", abstract = "In everyday life, people often depend on their theory of mind, i.e., their ability to reason about unobservable mental content of others to understand, explain, and predict their behaviour. Many agent-based models have been designed to develop computational theory of mind and analyze its effectiveness in various tasks and settings. However, most existing models are not generic (e.g., only applied in a given setting), not feasible (e.g., require too much information to be processed), or not human-inspired (e.g., do not capture the behavioral heuristics of humans). This hinders their applicability in many settings. Accordingly, we propose a new computational theory of mind, which captures the human decision heuristics of reasoning by abstracting individual beliefs about others. We specifically study computational affinity and show how it can be used in tandem with theory of mind reasoning when designing agent models for human-agent negotiation. We perform two-agent simulations to analyze the role of affinity in getting to agreements when there is a bound on the time to be spent for negotiating. Our results suggest that modeling affinity can ease the negotiation process by decreasing the number of rounds needed for an agreement as well as yield a higher benefit for agents with theory of mind reasoning." } - M. Michelini, A. Haret, and D. Grossi, “Group Wisdom at a Price: Jury Theorems with Costly Information,” in Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22, 2022, p. 419–425. doi:10.24963/ijcai.2022/60
[BibTeX] [Abstract] [Download PDF]
We study epistemic voting on binary issues where voters are characterized by their competence, i.e., the probability of voting for the correct alternative, and can choose between two actions: voting or abstaining. In our setting voting involves the expenditure of some effort, which is required to achieve the appropriate level of competence, whereas abstention carries no effort. We model this scenario as a game and characterize its equilibria under several variations. Our results show that when agents are aware of everyone’s incentives, then the addition of effort may lead to Nash equilibria where wisdom of the crowds is lost. We further show that if agents’ awareness of each other is constrained by a social network, the topology of the network may actually mitigate this effect.
@inproceedings{michelini22group, author = {Michelini, Matteo and Haret, Adrian and Grossi, Davide}, booktitle = {Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, {IJCAI-22}}, date-added = {2022-12-14 12:31:41 +0100}, date-modified ={2022-12-14 12:32:24 +0100}, doi = {10.24963/ijcai.2022/60}, editor = {Lud De Raedt}, month = {7}, note = {Main Track}, pages = {419--425}, publisher = {International Joint Conferences on Artificial Intelligence Organization}, title = {Group Wisdom at a Price: Jury Theorems with Costly Information}, url = {https://doi.org/10.24963/ijcai.2022/60}, year = {2022}, bdsk-url-1 = {https://doi.org/10.24963/ijcai.2022/60}, abstract = { We study epistemic voting on binary issues where voters are characterized by their competence, i.e., the probability of voting for the correct alternative, and can choose between two actions: voting or abstaining. In our setting voting involves the expenditure of some effort, which is required to achieve the appropriate level of competence, whereas abstention carries no effort. We model this scenario as a game and characterize its equilibria under several variations. Our results show that when agents are aware of everyone's incentives, then the addition of effort may lead to Nash equilibria where wisdom of the crowds is lost. We further show that if agents' awareness of each other is constrained by a social network, the topology of the network may actually mitigate this effect. } } - M. Los, Z. Christoff, and D. Grossi, “Proportional Budget Allocations: Towards a Systematization,” in Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22, 2022, p. 398–404. doi:10.24963/ijcai.2022/57
[BibTeX] [Abstract] [Download PDF]
We contribute to the programme of lifting proportionality axioms from the multi-winner voting setting to participatory budgeting. We define novel proportionality axioms for participatory budgeting and test them on known proportionality-driven rules such as Phragmén and Rule X. We investigate logical implications among old and new axioms and provide a systematic overview of proportionality criteria in participatory budgeting.
@inproceedings{los22proportional, author = {Los, Maaike and Christoff, Zoé and Grossi, Davide}, booktitle = {Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, {IJCAI-22}}, date-added = {2022-12-14 12:30:46 +0100}, date-modified ={2022-12-14 12:32:58 +0100}, doi = {10.24963/ijcai.2022/57}, editor = {Lud De Raedt}, month = {7}, note = {Main Track}, pages = {398--404}, publisher = {International Joint Conferences on Artificial Intelligence Organization}, title = {Proportional Budget Allocations: Towards a Systematization}, url = {https://doi.org/10.24963/ijcai.2022/57}, year = {2022}, bdsk-url-1 = {https://doi.org/10.24963/ijcai.2022/57}, abstract = {We contribute to the programme of lifting proportionality axioms from the multi-winner voting setting to participatory budgeting. We define novel proportionality axioms for participatory budgeting and test them on known proportionality-driven rules such as Phragm{\'e}n and Rule X. We investigate logical implications among old and new axioms and provide a systematic overview of proportionality criteria in participatory budgeting. } } - Y. Zhang and D. Grossi, “Tracking Truth by Weighting Proxies in Liquid Democracy,” in 21st International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2022, Auckland, New Zealand, May 9-13, 2022, 2022, p. 1482–1490. doi:10.5555/3535850.3536015
[BibTeX] [Abstract] [Download PDF]
We study wisdom-of-the-crowd effects in liquid democracy on net- works where agents are allowed to apportion parts of their voting weight to different proxies. We show that in this setting–-unlike in the standard one where voting weight is delegated in full to only one proxy–-it becomes possible to construct delegation struc- tures that optimize the truth-tracking ability of the group. Focusing on group accuracy we contrast this centralized solution with the setting in which agents are free to choose their weighted delega- tions by greedily trying to maximize their own individual accuracy. While equilibria with weighted delegations may be as bad as with standard delegations, they are never worse and may sometimes be better. To gain further insights into this model we experimentally study quantal response delegation strategies on random networks. We observe that weighted delegations can lead, under specific con- ditions, to higher group accuracy than simple majority voting
@inproceedings{zhang22tracking, author = {Yuzhe Zhang and Davide Grossi}, bibsource = {dblp computer science bibliography, https://dblp.org}, biburl = {https://dblp.org/rec/conf/atal/ZhangG22.bib}, booktitle = {21st International Conference on Autonomous Agents and Multiagent Systems, {AAMAS} 2022, Auckland, New Zealand, May 9-13, 2022}, date-added = {2022-12-14 12:29:16 +0100}, date-modified ={2022-12-14 12:33:14 +0100}, doi = {10.5555/3535850.3536015}, editor = {Piotr Faliszewski and Viviana Mascardi and Catherine Pelachaud and Matthew E. Taylor}, pages = {1482--1490}, publisher = {International Foundation for Autonomous Agents and Multiagent Systems {(IFAAMAS)}}, timestamp = {Mon, 18 Jul 2022 17:13:00 +0200}, title = {Tracking Truth by Weighting Proxies in Liquid Democracy}, url = {https://www.ifaamas.org/Proceedings/aamas2022/pdfs/p1482.pdf}, year = {2022}, abstract = {We study wisdom-of-the-crowd effects in liquid democracy on net- works where agents are allowed to apportion parts of their voting weight to different proxies. We show that in this setting---unlike in the standard one where voting weight is delegated in full to only one proxy---it becomes possible to construct delegation struc- tures that optimize the truth-tracking ability of the group. Focusing on group accuracy we contrast this centralized solution with the setting in which agents are free to choose their weighted delega- tions by greedily trying to maximize their own individual accuracy. While equilibria with weighted delegations may be as bad as with standard delegations, they are never worse and may sometimes be better. To gain further insights into this model we experimentally study quantal response delegation strategies on random networks. We observe that weighted delegations can lead, under specific con- ditions, to higher group accuracy than simple majority voting} } - D. Grossi, W. van der Hoek, and L. B. Kuijer, “Reasoning about General Preference Relations,” Artif. Intell., vol. 313, iss. C, 2022. doi:10.1016/j.artint.2022.103793
[BibTeX] [Abstract] [Download PDF]
Preference relations are at the heart of many fundamental concepts in artificial intelligence, ranging from utility comparisons, to defeat among strategies and relative plausibility among states, just to mention a few. Reasoning about such relations has been the object of extensive research and a wealth of formalisms exist to express and reason about them. One such formalism is conditional logic, which focuses on reasoning about the “best” alternatives according to a given preference relation. A “best” alternative is normally interpreted as an alternative that is either maximal (no other alternative is preferred to it) or optimal (it is at least as preferred as all other alternatives). And the preference relation is normally assumed to satisfy strong requirements (typically transitivity and some kind of well-foundedness assumption). Here, we generalize this existing literature in two ways. Firstly, in addition to maximality and optimality, we consider two other interpretations of “best”, which we call unmatchedness and acceptability. Secondly, we do not inherently require the preference relation to satisfy any constraints. Instead, we allow the relation to satisfy any combination of transitivity, totality and anti-symmetry. This allows us to model a wide range of situations, including cases where the lack of constraints stems from a modeled agent being irrational (for example, an agent might have preferences that are neither transitive nor total nor anti-symmetric) or from the interaction of perfectly rational agents (for example, a defeat relation among strategies in a game might be anti-symmetric but not total or transitive). For each interpretation of “best” (maximal, optimal, unmatched or acceptable) and each combination of constraints (transitivity, totality and/or anti-symmetry), we study the sets of valid inferences. Specifically, in all but one case we introduce a sound and strongly complete axiomatization, and in the one remaining case we show that no such axiomatization exists.
@article{grossi22reasoning, address = {GBR}, author = {Grossi, Davide and van der Hoek, Wiebe and Kuijer, Louwe B.}, date-added = {2022-12-14 12:27:55 +0100}, date-modified ={2022-12-14 12:33:06 +0100}, doi = {10.1016/j.artint.2022.103793}, issn = {0004-3702}, issue_date = {Dec 2022}, journal = {Artif. Intell.}, keywords = {Preference relations, Conditional logic}, month = {nov}, number = {C}, numpages = {44}, publisher = {Elsevier Science Publishers Ltd.}, title = {Reasoning about General Preference Relations}, url = {https://doi.org/10.1016/j.artint.2022.103793}, volume = {313}, year = {2022}, bdsk-url-1 = {https://doi.org/10.1016/j.artint.2022.103793}, abstract = {Preference relations are at the heart of many fundamental concepts in artificial intelligence, ranging from utility comparisons, to defeat among strategies and relative plausibility among states, just to mention a few. Reasoning about such relations has been the object of extensive research and a wealth of formalisms exist to express and reason about them. One such formalism is conditional logic, which focuses on reasoning about the ``best'' alternatives according to a given preference relation. A ``best'' alternative is normally interpreted as an alternative that is either maximal (no other alternative is preferred to it) or optimal (it is at least as preferred as all other alternatives). And the preference relation is normally assumed to satisfy strong requirements (typically transitivity and some kind of well-foundedness assumption). Here, we generalize this existing literature in two ways. Firstly, in addition to maximality and optimality, we consider two other interpretations of ``best'', which we call unmatchedness and acceptability. Secondly, we do not inherently require the preference relation to satisfy any constraints. Instead, we allow the relation to satisfy any combination of transitivity, totality and anti-symmetry. This allows us to model a wide range of situations, including cases where the lack of constraints stems from a modeled agent being irrational (for example, an agent might have preferences that are neither transitive nor total nor anti-symmetric) or from the interaction of perfectly rational agents (for example, a defeat relation among strategies in a game might be anti-symmetric but not total or transitive). For each interpretation of ``best'' (maximal, optimal, unmatched or acceptable) and each combination of constraints (transitivity, totality and/or anti-symmetry), we study the sets of valid inferences. Specifically, in all but one case we introduce a sound and strongly complete axiomatization, and in the one remaining case we show that no such axiomatization exists.} } - A. Onnes, Monitoring AI Systems: A Problem Analysis, Framework and OutlookIOS press, 2022.
[BibTeX] [Abstract] [Download PDF]
Knowledge-based systems have been used to monitor machines and processes in the real world. In this paper we propose the use of knowledge-based systems to monitor other AI systems in operation. We motivate and provide a problem analysis of this novel setting and subsequently propose a framework that allows for structuring future research related to this setting. Several directions for further research are also discussed.
@proceedings{Onnes2022, author = {Onnes, Annet}, title = {Monitoring AI Systems: A Problem Analysis, Framework and Outlook}, year = {2022}, booktitle = {HHAI 2022: Augmenting Human Intellect}, editor = {Schlobach, Stefan and Pérez-Ortiz, María and Tielman, Myrthe}, volume = {354}, series = {Frontiers in Artificial Intelligence and Applications}, publisher = {IOS press}, URL = {https://ebooks.iospress.nl/volumearticle/60870}, Abstract = {Knowledge-based systems have been used to monitor machines and processes in the real world. In this paper we propose the use of knowledge-based systems to monitor other AI systems in operation. We motivate and provide a problem analysis of this novel setting and subsequently propose a framework that allows for structuring future research related to this setting. Several directions for further research are also discussed.} } - K. Deja, A. Kuzina, T. Trzciński, and J. M. Tomczak, “On Analyzing Generative and Denoising Capabilities of Diffusion-based Deep Generative Models,” 36th Conference on Neural Information Processing Systems (NeurIPS 2022), 2022.
[BibTeX] [Abstract] [Download PDF]
Diffusion-based Deep Generative Models (DDGMs) offer state-of-the-art performance in generative modeling. Their main strength comes from their unique setup in which a model (the backward diffusion process) is trained to reverse the forward diffusion process, which gradually adds noise to the input signal. Although DDGMs are well studied, it is still unclear how the small amount of noise is transformed during the backward diffusion process. Here, we focus on analyzing this problem to gain more insight into the behavior of DDGMs and their denoising and generative capabilities. We observe a fluid transition point that changes the functionality of the backward diffusion process from generating a (corrupted) image from noise to denoising the corrupted image to the final sample. Based on this observation, we postulate to divide a DDGM into two parts: a denoiser and a generator. The denoiser could be parameterized by a denoising auto-encoder, while the generator is a diffusion-based model with its own set of parameters. We experimentally validate our proposition, showing its pros and cons.
@article{deja2022analyzing, title = {On Analyzing Generative and Denoising Capabilities of Diffusion-based Deep Generative Models}, author = {Deja, Kamil and Kuzina, Anna and Trzci{\'n}ski, Tomasz and Tomczak, Jakub M}, journal = {36th Conference on Neural Information Processing Systems (NeurIPS 2022)}, year = {2022}, url = {https://arxiv.org/abs/2206.00070}, abstract = {Diffusion-based Deep Generative Models (DDGMs) offer state-of-the-art performance in generative modeling. Their main strength comes from their unique setup in which a model (the backward diffusion process) is trained to reverse the forward diffusion process, which gradually adds noise to the input signal. Although DDGMs are well studied, it is still unclear how the small amount of noise is transformed during the backward diffusion process. Here, we focus on analyzing this problem to gain more insight into the behavior of DDGMs and their denoising and generative capabilities. We observe a fluid transition point that changes the functionality of the backward diffusion process from generating a (corrupted) image from noise to denoising the corrupted image to the final sample. Based on this observation, we postulate to divide a DDGM into two parts: a denoiser and a generator. The denoiser could be parameterized by a denoising auto-encoder, while the generator is a diffusion-based model with its own set of parameters. We experimentally validate our proposition, showing its pros and cons.} } - A. Kuzina, M. Welling, and J. M. Tomczak, “Alleviating Adversarial Attacks on Variational Autoencoders with MCMC,” 36th Conference on Neural Information Processing Systems (NeurIPS 2022), 2022.
[BibTeX] [Abstract] [Download PDF]
Variational autoencoders (VAEs) are latent variable models that can generate complex objects and provide meaningful latent representations. Moreover, they could be further used in downstream tasks such as classification. As previous work has shown, one can easily fool VAEs to produce unexpected latent representations and reconstructions for a visually slightly modified input. Here, we examine several objective functions for adversarial attack construction proposed previously and present a solution to alleviate the effect of these attacks. Our method utilizes the Markov Chain Monte Carlo (MCMC) technique in the inference step that we motivate with a theoretical analysis. Thus, we do not incorporate any extra costs during training, and the performance on non-attacked inputs is not decreased. We validate our approach on a variety of datasets (MNIST, Fashion MNIST, Color MNIST, CelebA) and VAE configurations (β -VAE, NVAE, β-TCVAE), and show that our approach consistently improves the model robustness to adversarial attacks.
@article{kuzina2022alleviating, title = {Alleviating Adversarial Attacks on Variational Autoencoders with MCMC}, author = {Kuzina, Anna and Welling, Max and Tomczak, Jakub M}, journal = {36th Conference on Neural Information Processing Systems (NeurIPS 2022)}, year = {2022}, url = {https://arxiv.org/abs/2203.09940}, abstract = {Variational autoencoders (VAEs) are latent variable models that can generate complex objects and provide meaningful latent representations. Moreover, they could be further used in downstream tasks such as classification. As previous work has shown, one can easily fool VAEs to produce unexpected latent representations and reconstructions for a visually slightly modified input. Here, we examine several objective functions for adversarial attack construction proposed previously and present a solution to alleviate the effect of these attacks. Our method utilizes the Markov Chain Monte Carlo (MCMC) technique in the inference step that we motivate with a theoretical analysis. Thus, we do not incorporate any extra costs during training, and the performance on non-attacked inputs is not decreased. We validate our approach on a variety of datasets (MNIST, Fashion MNIST, Color MNIST, CelebA) and VAE configurations (β -VAE, NVAE, β-TCVAE), and show that our approach consistently improves the model robustness to adversarial attacks.} } - S. Vadgama, J. M. Tomczak, and E. Bekkers, “Kendall Shape-VAE : Learning Shapes in a Generative Framework,” in NeurIPS 2022 Workshop on Symmetry and Geometry in Neural Representations, 2022.
[BibTeX] [Abstract] [Download PDF]
Learning an interpretable representation of data without supervision is an important precursor for the development of artificial intelligence. In this work, we introduce \textit{Kendall Shape}-VAE, a novel Variational Autoencoder framework for learning shapes as it disentangles the latent space by compressing information to simpler geometric symbols. In \textit{Kendall Shape}-VAE, we modify the Hyperspherical Variational Autoencoder such that it results in an exactly rotationally equivariant network using the notion of landmarks in the Kendall shape space. We show the exact equivariance of the model through experiments on rotated MNIST.
@inproceedings{vadgama2022kendall, title = {Kendall Shape-{VAE} : Learning Shapes in a Generative Framework}, author = {Sharvaree Vadgama and Jakub Mikolaj Tomczak and Erik J Bekkers}, booktitle = {NeurIPS 2022 Workshop on Symmetry and Geometry in Neural Representations}, year = {2022}, url = {https://openreview.net/forum?id=nzh4N6kdl2G}, abstract = {Learning an interpretable representation of data without supervision is an important precursor for the development of artificial intelligence. In this work, we introduce \textit{Kendall Shape}-VAE, a novel Variational Autoencoder framework for learning shapes as it disentangles the latent space by compressing information to simpler geometric symbols. In \textit{Kendall Shape}-VAE, we modify the Hyperspherical Variational Autoencoder such that it results in an exactly rotationally equivariant network using the notion of landmarks in the Kendall shape space. We show the exact equivariance of the model through experiments on rotated MNIST.} } - E. Liscio, A. E. Dondera, A. Geadau, C. M. Jonker, and P. K. Murukannaiah, “Cross-Domain Classification of Moral Values,” in Findings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics, Seattle, WA, USA, 2022, p. 2727–2745.
[BibTeX] [Abstract] [Download PDF]
Moral values influence how we interpret and act upon the information we receive. Identifying human moral values is essential for artificially intelligent agents to co-exist with humans. Recent progress in natural language processing allows the identification of moral values in textual discourse. However, domain-specific moral rhetoric poses challenges for transferring knowledge from one domain to another.We provide the first extensive investigation on the effects of cross-domain classification of moral values from text. We compare a state-of-the-art deep learning model (BERT) in seven domains and four cross-domain settings. We show that a value classifier can generalize and transfer knowledge to novel domains, but it can introduce catastrophic forgetting. We also highlight the typical classification errors in cross-domain value classification and compare the model predictions to the annotators agreement. Our results provide insights to computer and social scientists that seek to identify moral rhetoric specific to a domain of discourse.
@inproceedings{Liscio2022a, title = {{Cross-Domain Classification of Moral Values}}, year = {2022}, booktitle = {Findings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics}, series = {NAACL '22}, author = {Liscio, Enrico and Dondera, Alin E. and Geadau, Andrei and Jonker, Catholijn M. and Murukannaiah, Pradeep K.}, pages = {2727--2745}, publisher = {ACL}, address = {Seattle, WA, USA}, url = {https://aclanthology.org/2022.findings-naacl.209.pdf}, abstract = {Moral values influence how we interpret and act upon the information we receive. Identifying human moral values is essential for artificially intelligent agents to co-exist with humans. Recent progress in natural language processing allows the identification of moral values in textual discourse. However, domain-specific moral rhetoric poses challenges for transferring knowledge from one domain to another.We provide the first extensive investigation on the effects of cross-domain classification of moral values from text. We compare a state-of-the-art deep learning model (BERT) in seven domains and four cross-domain settings. We show that a value classifier can generalize and transfer knowledge to novel domains, but it can introduce catastrophic forgetting. We also highlight the typical classification errors in cross-domain value classification and compare the model predictions to the annotators agreement. Our results provide insights to computer and social scientists that seek to identify moral rhetoric specific to a domain of discourse.} } - L. C. Siebert, E. Liscio, P. K. Murukannaiah, L. Kaptein, S. L. Spruit, J. van den Hoven, and C. M. Jonker, “Estimating Value Preferences in a Hybrid Participatory System,” in HHAI2022: Augmenting Human Intellect, Amsterdam, the Netherlands, 2022, p. 114–127.
[BibTeX] [Abstract] [Download PDF]
We propose methods for an AI agent to estimate the value preferences of individuals in a hybrid participatory system, considering a setting where participants make choices and provide textual motivations for those choices. We focus on situations where there is a conflict between participants’ choices and motivations, and operationalize the philosophical stance that “valuing is deliberatively consequential. That is, if a user’s choice is based on a deliberation of value preferences, the value preferences can be observed in the motivation the user provides for the choice. Thus, we prioritize the value preferences estimated from motivations over the value preferences estimated from choices alone. We evaluate the proposed methods on a dataset of a large-scale survey on energy transition. The results show that explicitly addressing inconsistencies between choices and motivations improves the estimation of an individual’s value preferences. The proposed methods can be integrated in a hybrid participatory system, where artificial agents ought to estimate humans’ value preferences to pursue value alignment.
@inproceedings{Siebert2022, title = {{Estimating Value Preferences in a Hybrid Participatory System}}, year = {2022}, booktitle = {HHAI2022: Augmenting Human Intellect}, author = {Siebert, Luciano C. and Liscio, Enrico and Murukannaiah, Pradeep K. and Kaptein, Lionel and Spruit, Shannon L. and van den Hoven, Jeroen and Jonker, Catholijn M.}, pages = {114--127}, publisher = {IOS Press}, series = {HHAI '22}, address = {Amsterdam, the Netherlands}, url = {https://ebooks.iospress.nl/volumearticle/60861}, abstract = {We propose methods for an AI agent to estimate the value preferences of individuals in a hybrid participatory system, considering a setting where participants make choices and provide textual motivations for those choices. We focus on situations where there is a conflict between participants' choices and motivations, and operationalize the philosophical stance that “valuing is deliberatively consequential. That is, if a user's choice is based on a deliberation of value preferences, the value preferences can be observed in the motivation the user provides for the choice. Thus, we prioritize the value preferences estimated from motivations over the value preferences estimated from choices alone. We evaluate the proposed methods on a dataset of a large-scale survey on energy transition. The results show that explicitly addressing inconsistencies between choices and motivations improves the estimation of an individual's value preferences. The proposed methods can be integrated in a hybrid participatory system, where artificial agents ought to estimate humans' value preferences to pursue value alignment.} } - M. van der Meer, E. Liscio, C. M. Jonker, A. Plaat, P. Vossen, and P. K. Murukannaiah, “HyEnA: A Hybrid Method for Extracting Arguments from Opinions,” in HHAI2022: Augmenting Human Intellect, Amsterdam, the Netherlands, 2022, p. 17–31.
[BibTeX] [Abstract] [Download PDF]
The key arguments underlying a large and noisy set of opinions help understand the opinions quickly and accurately. Fully automated methods can extract arguments but (1) require large labeled datasets and (2) work well for known viewpoints, but not for novel points of view. We propose HyEnA, a hybrid (human + AI) method for extracting arguments from opinionated texts, combining the speed of automated processing with the understanding and reasoning capabilities of humans. We evaluate HyEnA on three feedback corpora. We find that, on the one hand, HyEnA achieves higher coverage and precision than a state-of-the-art automated method, when compared on a common set of diverse opinions, justifying the need for human insight. On the other hand, HyEnA requires less human effort and does not compromise quality compared to (fully manual) expert analysis, demonstrating the benefit of combining human and machine intelligence.
@inproceedings{vanderMeer2022, title = {{HyEnA: A Hybrid Method for Extracting Arguments from Opinions}}, year = {2022}, booktitle = {HHAI2022: Augmenting Human Intellect}, author = {van der Meer, Michiel and Liscio, Enrico and Jonker, Catholijn M. and Plaat, Aske and Vossen, Piek and Murukannaiah, Pradeep K.}, pages = {17--31}, publisher = {IOS Press}, series = {HHAI '22}, address = {Amsterdam, the Netherlands}, url = {https://ebooks.iospress.nl/volumearticle/60855}, abstract = {The key arguments underlying a large and noisy set of opinions help understand the opinions quickly and accurately. Fully automated methods can extract arguments but (1) require large labeled datasets and (2) work well for known viewpoints, but not for novel points of view. We propose HyEnA, a hybrid (human + AI) method for extracting arguments from opinionated texts, combining the speed of automated processing with the understanding and reasoning capabilities of humans. We evaluate HyEnA on three feedback corpora. We find that, on the one hand, HyEnA achieves higher coverage and precision than a state-of-the-art automated method, when compared on a common set of diverse opinions, justifying the need for human insight. On the other hand, HyEnA requires less human effort and does not compromise quality compared to (fully manual) expert analysis, demonstrating the benefit of combining human and machine intelligence.} } - R. Loftin and F. A. Oliehoek, “On the Impossibility of Learning to Cooperate with Adaptive Partner Strategies in Repeated Games,” in International Conference on Machine Learning, 2022, p. 14197–14209.
[BibTeX] [Abstract] [Download PDF]
Learning to cooperate with other agents is challenging when those agents also possess the ability to adapt to our own behavior. Practical and theoretical approaches to learning in cooperative settings typically assume that other agents’ behaviors are stationary, or else make very specific assumptions about other agents’ learning processes. The goal of this work is to understand whether we can reliably learn to cooperate with other agents without such restrictive assumptions, which are unlikely to hold in real-world applications. Our main contribution is a set of impossibility results, which show that no learning algorithm can reliably learn to cooperate with all possible adaptive partners in a repeated matrix game, even if that partner is guaranteed to cooperate with some stationary strategy. Motivated by these results, we then discuss potential alternative assumptions which capture the idea that an adaptive partner will only adapt rationally to our behavior.
@inproceedings{loftin2022impossibility, title = {On the Impossibility of Learning to Cooperate with Adaptive Partner Strategies in Repeated Games}, author = {Loftin, Robert and Oliehoek, Frans A}, booktitle = {International Conference on Machine Learning}, pages = {14197--14209}, year = {2022}, organization = {PMLR}, abstract = {Learning to cooperate with other agents is challenging when those agents also possess the ability to adapt to our own behavior. Practical and theoretical approaches to learning in cooperative settings typically assume that other agents' behaviors are stationary, or else make very specific assumptions about other agents' learning processes. The goal of this work is to understand whether we can reliably learn to cooperate with other agents without such restrictive assumptions, which are unlikely to hold in real-world applications. Our main contribution is a set of impossibility results, which show that no learning algorithm can reliably learn to cooperate with all possible adaptive partners in a repeated matrix game, even if that partner is guaranteed to cooperate with some stationary strategy. Motivated by these results, we then discuss potential alternative assumptions which capture the idea that an adaptive partner will only adapt rationally to our behavior.}, URL = {https://arxiv.org/abs/2206.10614} } - J. van Rhenen, C. Centeio Jorge, T. Matej Hrkalovic, and B. Dudzik, “Effects of Social Behaviours in Online Video Games on Team Trust,” in Extended Abstracts of the 2022 Annual Symposium on Computer-Human Interaction in Play, 2022, p. 159–165.
[BibTeX] [Abstract] [Download PDF]
In competitive multiplayer online video games, teamwork is of utmost importance, implying high levels of interdependence between the joint outcomes of players. When engaging in such interdependent interactions, humans rely on trust to facilitate coordination of their individual behaviours. However, online games often take place between teams of strangers, with individual members having little to no information about each other than what they observe throughout the interaction itself. A better understanding of the social behaviours that are used by players to form trust could not only facilitate richer gaming experiences, but could also lead to insights about team interactions. As such, this paper presents a first step towards understanding how and which types of in-game behaviour relate to trust formation. In particular, we investigate a)which in-game behaviour were relevant for trust formation (first part of the study) and b) how they relate to the reported player’s trust in their teammates (the second part of the study). The first part consisted of interviews with League of Legends players in order to create a taxonomy of in-game behaviours relevant for trust formation. As for the second part, we ran a small-scale pilot study where participants played the game and then answered a questionnaire to measure their trust in their teammates. Our preliminary results present a taxonomy of in-game behaviours which can be used to annotate the games regarding trust behaviours. Based on the pilot study, the list of behaviours could be extended as to improve the results. These findings can be used to research the role of trust formation in teamwork
@inproceedings{van2022effects, title={Effects of Social Behaviours in Online Video Games on Team Trust}, author={van Rhenen, Jan-Willem and Centeio Jorge, Carolina and Matej Hrkalovic, Tiffany and Dudzik, Bernd}, booktitle={Extended Abstracts of the 2022 Annual Symposium on Computer-Human Interaction in Play}, pages={159--165}, year={2022}, URL={https://pure.tudelft.nl/admin/files/146989190/vanRhenen2021_author.pdf}, Abstract={In competitive multiplayer online video games, teamwork is of utmost importance, implying high levels of interdependence between the joint outcomes of players. When engaging in such interdependent interactions, humans rely on trust to facilitate coordination of their individual behaviours. However, online games often take place between teams of strangers, with individual members having little to no information about each other than what they observe throughout the interaction itself. A better understanding of the social behaviours that are used by players to form trust could not only facilitate richer gaming experiences, but could also lead to insights about team interactions. As such, this paper presents a first step towards understanding how and which types of in-game behaviour relate to trust formation. In particular, we investigate a)which in-game behaviour were relevant for trust formation (first part of the study) and b) how they relate to the reported player’s trust in their teammates (the second part of the study). The first part consisted of interviews with League of Legends players in order to create a taxonomy of in-game behaviours relevant for trust formation. As for the second part, we ran a small-scale pilot study where participants played the game and then answered a questionnaire to measure their trust in their teammates. Our preliminary results present a taxonomy of in-game behaviours which can be used to annotate the games regarding trust behaviours. Based on the pilot study, the list of behaviours could be extended as to improve the results. These findings can be used to research the role of trust formation in teamwork} } - T. Matej Hrkalovic, “Designing Hybrid Intelligence Techniques for Facilitating Collaboration Informed by Social Science,” in Proceedings of the 2022 International Conference on Multimodal Interaction, 2022, p. 679–684.
[BibTeX] [Abstract] [Download PDF]
Designing (socially) intelligent systems for facilitating collaborations in human-human and human-AI teams will require them to have a basic understanding of principles underlying social decision- making. Partner selection – the ability to identify and select suitable partners for collaborative relationships – is one relevant component of social intelligence and an important ingredient for successful relationship management. In everyday life, decision to engage in joint undertakings are often based on impressions made during social interactions with potential partners. These impressions, and consequently, partner selection are informed by (non)-verbal behavioral cues. Despite its importance, research investigating how these impressions and partner selection decisions unfold in naturalistic settings seem to be lacking. Thus, in this paper, we present a project focused on understanding, predicting and modeling partner selection and understanding its relationship with human impressions in semi- naturalistic settings, such as social interactions, with the aim of informing future designing approaches of (hybrid) intelligence system that can understand, predict and aid in initiating and facilitating (current and future) collaborations.
@inproceedings{matej2022designing, title={Designing Hybrid Intelligence Techniques for Facilitating Collaboration Informed by Social Science}, author={Matej Hrkalovic, Tiffany}, booktitle={Proceedings of the 2022 International Conference on Multimodal Interaction}, pages={679--684}, year={2022}, URL={https://research.vu.nl/en/publications/designing-hybrid-intelligence-techniques-for-facilitating-collabo}, Abstract={Designing (socially) intelligent systems for facilitating collaborations in human-human and human-AI teams will require them to have a basic understanding of principles underlying social decision- making. Partner selection - the ability to identify and select suitable partners for collaborative relationships - is one relevant component of social intelligence and an important ingredient for successful relationship management. In everyday life, decision to engage in joint undertakings are often based on impressions made during social interactions with potential partners. These impressions, and consequently, partner selection are informed by (non)-verbal behavioral cues. Despite its importance, research investigating how these impressions and partner selection decisions unfold in naturalistic settings seem to be lacking. Thus, in this paper, we present a project focused on understanding, predicting and modeling partner selection and understanding its relationship with human impressions in semi- naturalistic settings, such as social interactions, with the aim of informing future designing approaches of (hybrid) intelligence system that can understand, predict and aid in initiating and facilitating (current and future) collaborations.} } - U. Khurana, I. Vermeulen, E. Nalisnick, M. Van Noorloos, and A. Fokkens, “Hate Speech Criteria: A Modular Approach to Task-Specific Hate Speech Definitions,” in Proceedings of the Sixth Workshop on Online Abuse and Harms (WOAH), Seattle, Washington (Hybrid), 2022, p. 176–191. doi:10.18653/v1/2022.woah-1.17
[BibTeX] [Abstract] [Download PDF]
The subjectivity of automatic hate speech detection makes it a complex task, reflected in different and incomplete definitions in NLP. We present hate speech criteria, developed with insights from a law and social science expert, that help researchers create more explicit definitions and annotation guidelines on five aspects: (1) target groups and (2) dominance, (3) perpetrator characteristics, (4) explicit presence of negative interactions, and the (5) type of consequences/effects. Definitions can be structured so that they cover a more broad or more narrow phenomenon and conscious choices can be made on specifying criteria or leaving them open. We argue that the goal and exact task developers have in mind should determine how the scope of hate speech is defined. We provide an overview of the properties of datasets from hatespeechdata.com that may help select the most suitable dataset for a specific scenario.
@inproceedings{khurana-etal-2022-hate, title = "Hate Speech Criteria: A Modular Approach to Task-Specific Hate Speech Definitions", author = "Khurana, Urja and Vermeulen, Ivar and Nalisnick, Eric and Van Noorloos, Marloes and Fokkens, Antske", booktitle = "Proceedings of the Sixth Workshop on Online Abuse and Harms (WOAH)", month = "jul", year = "2022", address = "Seattle, Washington (Hybrid)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2022.woah-1.17", doi = "10.18653/v1/2022.woah-1.17", pages = "176--191", abstract = "The subjectivity of automatic hate speech detection makes it a complex task, reflected in different and incomplete definitions in NLP. We present hate speech criteria, developed with insights from a law and social science expert, that help researchers create more explicit definitions and annotation guidelines on five aspects: (1) target groups and (2) dominance, (3) perpetrator characteristics, (4) explicit presence of negative interactions, and the (5) type of consequences/effects. Definitions can be structured so that they cover a more broad or more narrow phenomenon and conscious choices can be made on specifying criteria or leaving them open. We argue that the goal and exact task developers have in mind should determine how the scope of hate speech is defined. We provide an overview of the properties of datasets from hatespeechdata.com that may help select the most suitable dataset for a specific scenario.", }
2021
- M. van Bekkum, M. de Boer, F. van Harmelen, A. M. -, and A. Teije, “Modular design patterns for hybrid learning and reasoning systems,” Appl. Intell., vol. 51, iss. 9, p. 6528–6546, 2021. doi:10.1007/s10489-021-02394-3
[BibTeX] [Abstract] [Download PDF]
The unification of statistical (data-driven) and symbolic (knowledge-driven) methods is widely recognised as one of the key challenges of modern AI. Recent years have seen large number of publications on such hybrid neuro-symbolic AI systems. That rapidly growing literature is highly diverse and mostly empirical, and is lacking a unifying view of the large variety of these hybrid systems. In this paper we analyse a large body of recent literature and we propose a set of modular design patterns for such hybrid, neuro-symbolic systems. We are able to describe the architecture of a very large number of hybrid systems by composing only a small set of elementary patterns as building blocks. The main contributions of this paper are: 1) a taxonomically organised vocabulary to describe both processes and data structures used in hybrid systems; 2) a set of 15+ design patterns for hybrid AI systems, organised in a set of elementary patterns and a set of compositional patterns; 3) an application of these design patterns in two realistic use-cases for hybrid AI systems. Our patterns reveal similarities between systems that were not recognised until now. Finally, our design patterns extend and refine Kautz’ earlier attempt at categorising neuro-symbolic architectures.
@article{DBLP:journals/apin/BekkumBHMT21, author = {Michael van Bekkum and Maaike de Boer and Frank van Harmelen and Andr{\'{e}} Meyer{-}Vitali and Annette ten Teije}, title = {Modular design patterns for hybrid learning and reasoning systems}, journal = {Appl. Intell.}, volume = {51}, number = {9}, pages = {6528--6546}, year = {2021}, url = {https://doi.org/10.1007/s10489-021-02394-3}, doi = {10.1007/s10489-021-02394-3}, timestamp = {Wed, 01 Sep 2021 12:45:13 +0200}, biburl = {https://dblp.org/rec/journals/apin/BekkumBHMT21.bib}, url = {https://link.springer.com/article/10.1007/s10489-021-02394-3 }, abstract = {The unification of statistical (data-driven) and symbolic (knowledge-driven) methods is widely recognised as one of the key challenges of modern AI. Recent years have seen large number of publications on such hybrid neuro-symbolic AI systems. That rapidly growing literature is highly diverse and mostly empirical, and is lacking a unifying view of the large variety of these hybrid systems. In this paper we analyse a large body of recent literature and we propose a set of modular design patterns for such hybrid, neuro-symbolic systems. We are able to describe the architecture of a very large number of hybrid systems by composing only a small set of elementary patterns as building blocks. The main contributions of this paper are: 1) a taxonomically organised vocabulary to describe both processes and data structures used in hybrid systems; 2) a set of 15+ design patterns for hybrid AI systems, organised in a set of elementary patterns and a set of compositional patterns; 3) an application of these design patterns in two realistic use-cases for hybrid AI systems. Our patterns reveal similarities between systems that were not recognised until now. Finally, our design patterns extend and refine Kautz' earlier attempt at categorising neuro-symbolic architectures.} } - A. Kuzina, M. Welling, and J. M. Tomczak, “Diagnosing Vulnerability of Variational Auto-Encoders to Adversarial Attacks,” in ICLR 2021 Workshop on Robust and Reliable Machine Learning in the Real World, 2021.
[BibTeX] [Abstract] [Download PDF]
In this work, we explore adversarial attacks on the Variational Autoencoders (VAE). We show how to modify data point to obtain a prescribed latent code (supervised attack) or just get a drastically different code (unsupervised attack). We examine the influence of model modifications ($\beta$-VAE, NVAE) on the robustness of VAEs and suggest metrics to quantify it.
@inproceedings{kuzina2021diagnosing, title = {Diagnosing Vulnerability of Variational Auto-Encoders to Adversarial Attacks}, author = {Kuzina, Anna and Welling, Max and Tomczak, Jakub M}, year = {2021}, booktitle = {ICLR 2021 Workshop on Robust and Reliable Machine Learning in the Real World}, url = {https://arxiv.org/pdf/2103.06701.pdf}, abstract = {In this work, we explore adversarial attacks on the Variational Autoencoders (VAE). We show how to modify data point to obtain a prescribed latent code (supervised attack) or just get a drastically different code (unsupervised attack). We examine the influence of model modifications ($\beta$-VAE, NVAE) on the robustness of VAEs and suggest metrics to quantify it.} } - H. Zheng and B. Verheij, “Rules, cases and arguments in artificial intelligence and law,” in Research Handbook on Big Data Law, R. Vogl, Ed., Edgar Elgar Publishing, 2021, pp. 373-387.
[BibTeX] [Abstract] [Download PDF]
Artificial intelligence and law is an interdisciplinary field of research that dates back at least to the 1970s, with academic conferences starting in the 1980s. In the field, complex problems are addressed about the computational modeling and automated support of legal reasoning and argumentation. Scholars have different backgrounds, and progress is driven by insights from lawyers, judges, computer scientists, philosophers and others. The community investigates and develops artificial intelligence techniques applicable in the legal domain, in order to enhance access to law for citizens and to support the efficiency and quality of work in the legal domain, aiming to promote a just society. Integral to the legal domain, legal reasoning and its structure and process have gained much attention in AI & Law research. Such research is today especially relevant, since in these days of big data and widespread use of algorithms, there is a need in AI to connect knowledge-based and data-driven AI techniques in order to arrive at a social, explainable and responsible AI. By considering knowledge in the form of rules and data in the form of cases connected by arguments, the field of AI & Law contributes relevant representations and algorithms for handling a combination of knowledge and data. In this chapter, as an entry point into the literature on AI & Law, three major styles of modeling legal reasoning are studied: rule-based reasoning, case-based reasoning and argument-based reasoning, which are the focus of this chapter. We describe selected key ideas, leaving out formal detail. As we will see, these styles of modeling legal reasoning are related, and there is much research investigating relations. We use the example domain of Dutch tort law (Section 2) to illustrate these three major styles, which are then more fully explained (Sections 3 to 5)
@InCollection{Zheng:2021, author = {H. Zheng and B. Verheij}, title = {Rules, cases and arguments in artificial intelligence and law}, booktitle = {Research Handbook on Big Data Law}, publisher = {Edgar Elgar Publishing}, editor = {R Vogl}, year = 2021, url = {https://www.ai.rug.nl/~verheij/publications/handbook2021.htm}, pages = {373-387}, abstract = {Artificial intelligence and law is an interdisciplinary field of research that dates back at least to the 1970s, with academic conferences starting in the 1980s. In the field, complex problems are addressed about the computational modeling and automated support of legal reasoning and argumentation. Scholars have different backgrounds, and progress is driven by insights from lawyers, judges, computer scientists, philosophers and others. The community investigates and develops artificial intelligence techniques applicable in the legal domain, in order to enhance access to law for citizens and to support the efficiency and quality of work in the legal domain, aiming to promote a just society. Integral to the legal domain, legal reasoning and its structure and process have gained much attention in AI & Law research. Such research is today especially relevant, since in these days of big data and widespread use of algorithms, there is a need in AI to connect knowledge-based and data-driven AI techniques in order to arrive at a social, explainable and responsible AI. By considering knowledge in the form of rules and data in the form of cases connected by arguments, the field of AI & Law contributes relevant representations and algorithms for handling a combination of knowledge and data. In this chapter, as an entry point into the literature on AI & Law, three major styles of modeling legal reasoning are studied: rule-based reasoning, case-based reasoning and argument-based reasoning, which are the focus of this chapter. We describe selected key ideas, leaving out formal detail. As we will see, these styles of modeling legal reasoning are related, and there is much research investigating relations. We use the example domain of Dutch tort law (Section 2) to illustrate these three major styles, which are then more fully explained (Sections 3 to 5)} } - C. A. Kurtan and P. i, “Assisting humans in privacy management: an agent-based approach,” Autonomous Agents and Multi-Agent Systems, vol. 35, iss. 7, 2021. doi:https://doi.org/10.1007/s10458-020-09488-1
[BibTeX] [Abstract] [Download PDF]
Image sharing is a service offered by many online social networks. In order to preserve privacy of images, users need to think through and specify a privacy setting for each image that they upload. This is difficult for two main reasons: first, research shows that many times users do not know their own privacy preferences, but only become aware of them over time. Second, even when users know their privacy preferences, editing these privacy settings is cumbersome and requires too much effort, interfering with the quick sharing behavior expected on an online social network. Accordingly, this paper proposes a privacy recommendation model for images using tags and an agent that implements this, namely pelte. Each user agent makes use of the privacy settings that its user have set for previous images to predict automatically the privacy setting for an image that is uploaded to be shared. When in doubt, the agent analyzes the sharing behavior of other users in the user’s network to be able to recommend to its user about what should be considered as private. Contrary to existing approaches that assume all the images are available to a centralized model, pelte is compatible to distributed environments since each agent accesses only the privacy settings of the images that the agent owner has shared or those that have been shared with the user. Our simulations on a real-life dataset shows that pelte can accurately predict privacy settings even when a user has shared a few images with others, the images have only a few tags or the user’s friends have varying privacy preferences.
@Article{kurtan-yolum-21, author = {A. Can Kurtan and P{\i}nar Yolum}, title = {Assisting humans in privacy management: an agent-based approach}, journal = {Autonomous Agents and Multi-Agent Systems}, year = {2021}, volume = {35}, number = {7}, abstract = {Image sharing is a service offered by many online social networks. In order to preserve privacy of images, users need to think through and specify a privacy setting for each image that they upload. This is difficult for two main reasons: first, research shows that many times users do not know their own privacy preferences, but only become aware of them over time. Second, even when users know their privacy preferences, editing these privacy settings is cumbersome and requires too much effort, interfering with the quick sharing behavior expected on an online social network. Accordingly, this paper proposes a privacy recommendation model for images using tags and an agent that implements this, namely pelte. Each user agent makes use of the privacy settings that its user have set for previous images to predict automatically the privacy setting for an image that is uploaded to be shared. When in doubt, the agent analyzes the sharing behavior of other users in the user's network to be able to recommend to its user about what should be considered as private. Contrary to existing approaches that assume all the images are available to a centralized model, pelte is compatible to distributed environments since each agent accesses only the privacy settings of the images that the agent owner has shared or those that have been shared with the user. Our simulations on a real-life dataset shows that pelte can accurately predict privacy settings even when a user has shared a few images with others, the images have only a few tags or the user's friends have varying privacy preferences.}, url = {https://link.springer.com/article/10.1007/s10458-020-09488-1}, doi = {https://doi.org/10.1007/s10458-020-09488-1} } - E. Liscio, M. van der Meer, C. M. Jonker, and P. K. Murukannaiah, “A Collaborative Platform for Identifying Context-Specific Values,” in Proceedings of the 20th International Conference on Autonomous Agents and Multiagent Systems, Online, 2021, p. 1773–1775.
[BibTeX] [Abstract] [Download PDF]
Value alignment is a crucial aspect of ethical multiagent systems. An important step toward value alignment is identifying values specific to an application context. However, identifying context-specific values is complex and cognitively demanding. To support this process, we develop a methodology and a collaborative web platform that employs AI techniques. We describe this platform, highlighting its intuitive design and implementation.
@inproceedings{Liscio2021a, title = {{A Collaborative Platform for Identifying Context-Specific Values}}, year = {2021}, booktitle = {Proceedings of the 20th International Conference on Autonomous Agents and Multiagent Systems}, series = {AAMAS '21}, author = {Liscio, Enrico and van der Meer, Michiel and Jonker, Catholijn M. and Murukannaiah, Pradeep K.}, pages = {1773--1775}, publisher = {IFAAMAS}, address = {Online}, url = {https://www.ifaamas.org/Proceedings/aamas2021/pdfs/p1773.pdf}, abstract = {Value alignment is a crucial aspect of ethical multiagent systems. An important step toward value alignment is identifying values specific to an application context. However, identifying context-specific values is complex and cognitively demanding. To support this process, we develop a methodology and a collaborative web platform that employs AI techniques. We describe this platform, highlighting its intuitive design and implementation.} } - E. Liscio, M. van der Meer, L. C. Siebert, C. M. Jonker, N. Mouter, and P. K. Murukannaiah, “Axies: Identifying and Evaluating Context-Specific Values,” in Proc. of the 20th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2021), Online, 2021, p. 799–808.
[BibTeX] [Abstract] [Download PDF]
The pursuit of values drives human behavior and promotes cooperation. Existing research is focused on general (e.g., Schwartz) values that transcend contexts. However, context-specific values are necessary to (1) understand human decisions, and (2) engineer intelligent agents that can elicit human values and take value-aligned actions. We propose Axies, a hybrid (human and AI) methodology to identify context-specific values. Axies simplifies the abstract task of value identification as a guided value annotation process involving human annotators. Axies exploits the growing availability of value-laden text corpora and Natural Language Processing to assist the annotators in systematically identifying context-specific values. We evaluate Axies in a user study involving 60 subjects. In our study, six annotators generate value lists for two timely and important contexts: Covid-19 measures, and sustainable Energy. Then, two policy experts and 52 crowd workers evaluate Axies value lists. We find that Axies yields values that are context-specific, consistent across different annotators, and comprehensible to end users
@inproceedings{Liscio2021b, address = {Online}, author = {Liscio, Enrico and van der Meer, Michiel and Siebert, Luciano C. and Jonker, Catholijn M. and Mouter, Niek and Murukannaiah, Pradeep K.}, booktitle = {Proc. of the 20th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2021)}, keywords = {Context,Ethics,Natural Language Processing,Values,acm reference format,catholijn m,context,enrico liscio,ethics,jonker,luciano c,michiel van der meer,natural language processing,siebert,values}, pages = {799--808}, publisher = {IFAAMAS}, title = {{Axies: Identifying and Evaluating Context-Specific Values}}, year = {2021}, url = "https://ii.tudelft.nl/~pradeep/doc/Liscio-2021-AAMAS-Axies.pdf", abstract = "The pursuit of values drives human behavior and promotes cooperation. Existing research is focused on general (e.g., Schwartz) values that transcend contexts. However, context-specific values are necessary to (1) understand human decisions, and (2) engineer intelligent agents that can elicit human values and take value-aligned actions. We propose Axies, a hybrid (human and AI) methodology to identify context-specific values. Axies simplifies the abstract task of value identification as a guided value annotation process involving human annotators. Axies exploits the growing availability of value-laden text corpora and Natural Language Processing to assist the annotators in systematically identifying context-specific values. We evaluate Axies in a user study involving 60 subjects. In our study, six annotators generate value lists for two timely and important contexts: Covid-19 measures, and sustainable Energy. Then, two policy experts and 52 crowd workers evaluate Axies value lists. We find that Axies yields values that are context-specific, consistent across different annotators, and comprehensible to end users" } - K. Miras, J. Cuijpers, B. Gülhan, and A. Eiben, “The Impact of Early-death on Phenotypically Plastic Robots that Evolve in Changing Environments,” in ALIFE 2021: The 2021 Conference on Artificial Life, 2021.
[BibTeX] [Abstract] [Download PDF]
In this work, we evolve phenotypically plastic robots-robots that adapt their bodies and brains according to environmental conditions-in changing environments. In particular, we investigate how the possibility of death in early environmental conditions impacts evolvability and robot traits. Our results demonstrate that early-death improves the efficiency of the evolutionary process for the earlier environmental conditions. On the other hand, the possibility of early-death in the earlier environmental conditions results in a dramatic loss of performance in the latter environmental conditions.
@inproceedings{miras2021impact, title = {The Impact of Early-death on Phenotypically Plastic Robots that Evolve in Changing Environments}, author = {Miras, Karine and Cuijpers, Jim and Gülhan, Bahadir and Eiben, AE}, booktitle = {ALIFE 2021: The 2021 Conference on Artificial Life}, year = {2021}, organization = {MIT Press}, url = "https://direct.mit.edu/isal/proceedings-pdf/isal/33/25/1929813/isal_a_00371.pdf", abstract = "In this work, we evolve phenotypically plastic robots-robots that adapt their bodies and brains according to environmental conditions-in changing environments. In particular, we investigate how the possibility of death in early environmental conditions impacts evolvability and robot traits. Our results demonstrate that early-death improves the efficiency of the evolutionary process for the earlier environmental conditions. On the other hand, the possibility of early-death in the earlier environmental conditions results in a dramatic loss of performance in the latter environmental conditions." } - K. Miras, “Constrained by Design: Influence of Genetic Encodings on Evolved Traits of Robots,” Frontiers Robotics AI, vol. 8, p. 672379, 2021. doi:10.3389/frobt.2021.672379
[BibTeX] [Abstract] [Download PDF]
Genetic encodings and their particular properties are known to have a strong influence on the success of evolutionary systems. However, the literature has widely focused on studying the effects that encodings have on performance, i.e., fitness-oriented studies. Notably, this anchoring of the literature to performance is limiting, considering that performance provides bounded information about the behavior of a robot system. In this paper, we investigate how genetic encodings constrain the space of robot phenotypes and robot behavior. In summary, we demonstrate how two generative encodings of different nature lead to very different robots and discuss these differences. Our principal contributions are creating awareness about robot encoding biases, demonstrating how such biases affect evolved morphological, control, and behavioral traits, and finally scrutinizing the trade-offs among different biases.
@article{DBLP:journals/firai/Miras21, author = {Karine Miras}, title = {Constrained by Design: Influence of Genetic Encodings on Evolved Traits of Robots}, journal = {Frontiers Robotics {AI}}, volume = {8}, pages = {672379}, year = {2021}, url = {https://doi.org/10.3389/frobt.2021.672379}, doi = {10.3389/frobt.2021.672379}, abstract = "Genetic encodings and their particular properties are known to have a strong influence on the success of evolutionary systems. However, the literature has widely focused on studying the effects that encodings have on performance, i.e., fitness-oriented studies. Notably, this anchoring of the literature to performance is limiting, considering that performance provides bounded information about the behavior of a robot system. In this paper, we investigate how genetic encodings constrain the space of robot phenotypes and robot behavior. In summary, we demonstrate how two generative encodings of different nature lead to very different robots and discuss these differences. Our principal contributions are creating awareness about robot encoding biases, demonstrating how such biases affect evolved morphological, control, and behavioral traits, and finally scrutinizing the trade-offs among different biases." } - P. Manggala, H. H. Hoos, and E. Nalisnick, “Bayesian Regression from Multiple Sources of Weak Supervision,” in ICML 2021 Machine Learning for Data: Automated Creation, Privacy, Bias, 2021.
[BibTeX] [Abstract] [Download PDF]
We describe a Bayesian approach to weakly supervised regression. Our proposed framework propagates uncertainty from the weak supervision to an aggregated predictive distribution. We use a generalized Bayes procedure to account for the supervision being weak and therefore likely misspecified.
@inproceedings{manggala2021bayesianregression, title = {Bayesian Regression from Multiple Sources of Weak Supervision}, author = {Manggala, Putra and Hoos, Holger H. and Nalisnick, Eric}, year = {2021}, booktitle = {ICML 2021 Machine Learning for Data: Automated Creation, Privacy, Bias}, url = {https://pmangg.github.io/papers/brfmsows_mhn_ml4data_icml.pdf}, abstract = {We describe a Bayesian approach to weakly supervised regression. Our proposed framework propagates uncertainty from the weak supervision to an aggregated predictive distribution. We use a generalized Bayes procedure to account for the supervision being weak and therefore likely misspecified.} } - C. Steging, S. Renooij, and B. Verheij, “Discovering the rationale of decisions: towards a method for aligning learning and reasoning,” in ICAIL ’21: Eighteenth International Conference for Artificial Intelligence and Law, São Paulo Brazil, June 21 – 25, 2021, 2021, p. 235–239. doi:10.1145/3462757.3466059
[BibTeX] [Abstract] [Download PDF]
In AI and law, systems that are designed for decision support should be explainable when pursuing justice. In order for these systems to be fair and responsible, they should make correct decisions and make them using a sound and transparent rationale. In this paper, we introduce a knowledge-driven method for model-agnostic rationale evaluation using dedicated test cases, similar to unit-testing in professional software development. We apply this new quantitative human-in-the-loop method in a machine learning experiment aimed at extracting known knowledge structures from artificial datasets from a real-life legal setting. We show that our method allows us to analyze the rationale of black box machine learning systems by assessing which rationale elements are learned or not. Furthermore, we show that the rationale can be adjusted using tailor-made training data based on the results of the rationale evaluation.
@inproceedings{StegingICAIL21, author = {Cor Steging and Silja Renooij and Bart Verheij}, editor = {Juliano Maranh{\~{a}}o and Adam Zachary Wyner}, title = {Discovering the rationale of decisions: towards a method for aligning learning and reasoning}, booktitle = {{ICAIL} '21: Eighteenth International Conference for Artificial Intelligence and Law, S{\~{a}}o Paulo Brazil, June 21 - 25, 2021}, pages = {235--239}, publisher = {{ACM}}, year = {2021}, url = {https://doi.org/10.1145/3462757.3466059}, doi = {10.1145/3462757.3466059}, abstract = "In AI and law, systems that are designed for decision support should be explainable when pursuing justice. In order for these systems to be fair and responsible, they should make correct decisions and make them using a sound and transparent rationale. In this paper, we introduce a knowledge-driven method for model-agnostic rationale evaluation using dedicated test cases, similar to unit-testing in professional software development. We apply this new quantitative human-in-the-loop method in a machine learning experiment aimed at extracting known knowledge structures from artificial datasets from a real-life legal setting. We show that our method allows us to analyze the rationale of black box machine learning systems by assessing which rationale elements are learned or not. Furthermore, we show that the rationale can be adjusted using tailor-made training data based on the results of the rationale evaluation." } - C. Steging, S. Renooij, and B. Verheij, “Discovering the Rationale of Decisions: Experiments on Aligning Learning and Reasoning,” in 4th EXplainable AI in Law Workshop (XAILA 2021), 2021, p. 235–239.
[BibTeX] [Abstract] [Download PDF]
In AI and law, systems that are designed for decision support should be explainable when pursuing justice. In order for these systems to be fair and responsible, they should make correct decisions and make them using a sound and transparent rationale. In this paper, we introduce a knowledge-driven method for model-agnostic rationale evaluation using dedicated test cases, similar to unit-testing in professional software development. We apply this new method in a set of machine learning experiments aimed at extracting known knowledge structures from artificial datasets from fictional and non-fictional legal settings. We show that our method allows us to analyze the rationale of black-box machine learning systems by assessing which rationale elements are learned or not. Furthermore, we show that the rationale can be adjusted using tailor-made training data based on the results of the rationale evaluation.
@inproceedings{StegingXAILA21, author = {Cor Steging and Silja Renooij and Bart Verheij}, title = {Discovering the Rationale of Decisions: Experiments on Aligning Learning and Reasoning}, maintitle = {{ICAIL} '21: Eighteenth International Conference for Artificial Intelligence and Law, S{\~{a}}o Paulo Brazil, June 21 - 25, 2021}, booktitle = {4th EXplainable AI in Law Workshop (XAILA 2021) }, pages = {235--239}, publisher = {{ACM}}, year = {2021}, url = {https://arxiv.org/abs/2105.06758}, abstract = "In AI and law, systems that are designed for decision support should be explainable when pursuing justice. In order for these systems to be fair and responsible, they should make correct decisions and make them using a sound and transparent rationale. In this paper, we introduce a knowledge-driven method for model-agnostic rationale evaluation using dedicated test cases, similar to unit-testing in professional software development. We apply this new method in a set of machine learning experiments aimed at extracting known knowledge structures from artificial datasets from fictional and non-fictional legal settings. We show that our method allows us to analyze the rationale of black-box machine learning systems by assessing which rationale elements are learned or not. Furthermore, we show that the rationale can be adjusted using tailor-made training data based on the results of the rationale evaluation." } - U. Khurana, E. Nalisnick, and A. Fokkens, “How Emotionally Stable is ALBERT? Testing Robustness with Stochastic Weight Averaging on a Sentiment Analysis Task,” in Proceedings of the 2nd Workshop on Evaluation and Comparison of NLP Systems, Punta Cana, Dominican Republic, 2021, p. 16–31.
[BibTeX] [Abstract] [Download PDF]
Despite their success, modern language models are fragile. Even small changes in their training pipeline can lead to unexpected results. We study this phenomenon by examining the robustness of ALBERT (Lan et al., 2020) in combination with Stochastic Weight Averaging (SWA){–-}a cheap way of ensembling{–-}on a sentiment analysis task (SST-2). In particular, we analyze SWA{‘}s stability via CheckList criteria (Ribeiro et al., 2020), examining the agreement on errors made by models differing only in their random seed. We hypothesize that SWA is more stable because it ensembles model snapshots taken along the gradient descent trajectory. We quantify stability by comparing the models{‘} mistakes with Fleiss{‘} Kappa (Fleiss, 1971) and overlap ratio scores. We find that SWA reduces error rates in general; yet the models still suffer from their own distinct biases (according to CheckList).
@inproceedings{khurana-etal-2021-emotionally, title = "How Emotionally Stable is {ALBERT}? Testing Robustness with Stochastic Weight Averaging on a Sentiment Analysis Task", author = "Khurana, Urja and Nalisnick, Eric and Fokkens, Antske", booktitle = "Proceedings of the 2nd Workshop on Evaluation and Comparison of NLP Systems", month = nov, year = "2021", address = "Punta Cana, Dominican Republic", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.eval4nlp-1.3", pages = "16--31", abstract = "Despite their success, modern language models are fragile. Even small changes in their training pipeline can lead to unexpected results. We study this phenomenon by examining the robustness of ALBERT (Lan et al., 2020) in combination with Stochastic Weight Averaging (SWA){---}a cheap way of ensembling{---}on a sentiment analysis task (SST-2). In particular, we analyze SWA{'}s stability via CheckList criteria (Ribeiro et al., 2020), examining the agreement on errors made by models differing only in their random seed. We hypothesize that SWA is more stable because it ensembles model snapshots taken along the gradient descent trajectory. We quantify stability by comparing the models{'} mistakes with Fleiss{'} Kappa (Fleiss, 1971) and overlap ratio scores. We find that SWA reduces error rates in general; yet the models still suffer from their own distinct biases (according to CheckList).", } - M. B. Vessies, S. P. Vadgama, R. van de Leur, P. A. F. M. Doevendans, R. J. Hassink, E. Bekkers, and R. Es, “Interpretable ECG classification via a query-based latent space traversal (qLST),” CoRR, vol. abs/2111.07386, 2021.
[BibTeX] [Abstract] [Download PDF]
Electrocardiography (ECG) is an effective and non-invasive diagnostic tool that measures the electrical activity of the heart. Interpretation of ECG signals to detect various abnormalities is a challenging task that requires expertise. Recently, the use of deep neural networks for ECG classification to aid medical practitioners has become popular, but their black box nature hampers clinical implementation. Several saliency-based interpretability techniques have been proposed, but they only indicate the location of important features and not the actual features. We present a novel interpretability technique called qLST, a query-based latent space traversal technique that is able to provide explanations for any ECG classification model. With qLST, we train a neural network that learns to traverse in the latent space of a variational autoencoder trained on a large university hospital dataset with over 800,000 ECGs annotated for 28 diseases. We demonstrate through experiments that we can explain different black box classifiers by generating ECGs through these traversals.
@article{DBLP:journals/corr/abs-2111-07386, author = {Melle B. Vessies and Sharvaree P. Vadgama and Rutger R. van de Leur and Pieter A. F. M. Doevendans and Rutger J. Hassink and Erik Bekkers and Ren{\'{e}} van Es}, title = {Interpretable {ECG} classification via a query-based latent space traversal (qLST)}, journal = {CoRR}, volume = {abs/2111.07386}, year = {2021}, url = {https://arxiv.org/abs/2111.07386}, abstract = {Electrocardiography (ECG) is an effective and non-invasive diagnostic tool that measures the electrical activity of the heart. Interpretation of ECG signals to detect various abnormalities is a challenging task that requires expertise. Recently, the use of deep neural networks for ECG classification to aid medical practitioners has become popular, but their black box nature hampers clinical implementation. Several saliency-based interpretability techniques have been proposed, but they only indicate the location of important features and not the actual features. We present a novel interpretability technique called qLST, a query-based latent space traversal technique that is able to provide explanations for any ECG classification model. With qLST, we train a neural network that learns to traverse in the latent space of a variational autoencoder trained on a large university hospital dataset with over 800,000 ECGs annotated for 28 diseases. We demonstrate through experiments that we can explain different black box classifiers by generating ECGs through these traversals.} } - B. Dudzik and J. Broekens, “A Valid Self-Report is Never Late, Nor is it Early: On Considering the Right Temporal Distance for Assessing Emotional Experience,” , 1, 2021.
[BibTeX] [Abstract]
Developing computational models for automatic affect prediction requires valid self-reports about individuals’ emotional interpretations of stimuli. In this article, we highlight the important influence of the temporal distance between a stimulus event and the moment when its experience is reported on the provided information’s validity. This influence stems from the time-dependent and time-demanding nature of the involved cognitive processes. As such, reports can be collected too late: forgetting is a widely acknowledged challenge for accurate descriptions of past experience. For this reason, methods striving for assessment as early as possible have become increasingly popular. However, here we argue that collection may also occur too early: descriptions about very recent stimuli might be collected before emotional processing has fully converged. Based on these notions, we champion the existence of a temporal distance for each type of stimulus that maximizes the validity of self-reports–a” right” time. Consequently, we recommend future research to (1) consciously consider the potential influence of temporal distance on affective self-reports when planning data collection,(2) document the temporal distance of affective self-reports wherever possible as part of corpora for computational modelling, and finally (3) and explore the effect of temporal distance on self-reports across different types of stimuli.
@techreport{Dudzik2021, author = {Dudzik, Bernd and Broekens, Joost}, booktitle = {Momentary Emotion Elicitation and Capture Workshop (MEEC'21), May 9, 2021, Yokohama, Japan}, number = {1}, publisher = {Association for Computing Machinery}, title = {{A Valid Self-Report is Never Late, Nor is it Early: On Considering the Right Temporal Distance for Assessing Emotional Experience}}, volume = {1}, year = {2021}, abstract = {Developing computational models for automatic affect prediction requires valid self-reports about individuals’ emotional interpretations of stimuli. In this article, we highlight the important influence of the temporal distance between a stimulus event and the moment when its experience is reported on the provided information’s validity. This influence stems from the time-dependent and time-demanding nature of the involved cognitive processes. As such, reports can be collected too late: forgetting is a widely acknowledged challenge for accurate descriptions of past experience. For this reason, methods striving for assessment as early as possible have become increasingly popular. However, here we argue that collection may also occur too early: descriptions about very recent stimuli might be collected before emotional processing has fully converged. Based on these notions, we champion the existence of a temporal distance for each type of stimulus that maximizes the validity of self-reports–a" right" time. Consequently, we recommend future research to (1) consciously consider the potential influence of temporal distance on affective self-reports when planning data collection,(2) document the temporal distance of affective self-reports wherever possible as part of corpora for computational modelling, and finally (3) and explore the effect of temporal distance on self-reports across different types of stimuli.} } - B. Dudzik, S. Columbus, T. M. Hrkalovic, D. Balliet, and H. Hung, “Recognizing Perceived Interdependence in Face-to-Face Negotiations through Multimodal Analysis of Nonverbal Behavior,” in Proceedings of the 2021 International Conference on Multimodal Interaction, New York, NY, USA: Association for Computing Machinery, 2021, pp. 121-130. doi:10.1145/3462244.3479935
[BibTeX] [Abstract] [Download PDF]
Enabling computer-based applications to display intelligent behavior in complex social settings requires them to relate to important aspects of how humans experience and understand such situations. One crucial driver of peoples’ social behavior during an interaction is the interdependence they perceive, i.e., how the outcome of an interaction is determined by their own and others’ actions. According to psychological studies, both the nonverbal behavior displayed by Motivated by this, we present a series of experiments to automatically recognize interdependence perceptions in dyadic face-to-face negotiations using these sources. Concretely, our approach draws on a combination of features describing individuals’ Facial, Upper Body, and Vocal Behavior with state-of-the-art algorithms for multivariate time series classification. Our findings demonstrate that differences in some types of interdependence perceptions can be detected through the automatic analysis of nonverbal behaviors. We discuss implications for developing socially intelligent systems and opportunities for future research.
@inbook{10.1145/3462244.3479935, author = {Dudzik, Bernd and Columbus, Simon and Hrkalovic, Tiffany Matej and Balliet, Daniel and Hung, Hayley}, title = {Recognizing Perceived Interdependence in Face-to-Face Negotiations through Multimodal Analysis of Nonverbal Behavior}, year = {2021}, isbn = {9781450384810}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, booktitle = {Proceedings of the 2021 International Conference on Multimodal Interaction}, pages = {121-130}, numpages = {10}, doi = {10.1145/3462244.3479935}, url = {https://research.tudelft.nl/en/publications/recognizing-perceived-interdependence-in-face-to-face-negotiation}, abstract = {Enabling computer-based applications to display intelligent behavior in complex social settings requires them to relate to important aspects of how humans experience and understand such situations. One crucial driver of peoples' social behavior during an interaction is the interdependence they perceive, i.e., how the outcome of an interaction is determined by their own and others' actions. According to psychological studies, both the nonverbal behavior displayed by Motivated by this, we present a series of experiments to automatically recognize interdependence perceptions in dyadic face-to-face negotiations using these sources. Concretely, our approach draws on a combination of features describing individuals' Facial, Upper Body, and Vocal Behavior with state-of-the-art algorithms for multivariate time series classification. Our findings demonstrate that differences in some types of interdependence perceptions can be detected through the automatic analysis of nonverbal behaviors. We discuss implications for developing socially intelligent systems and opportunities for future research.} } - C. Steging, S. Renooij, and B. Verheij, “Rationale Discovery and Explainable AI,” in Legal Knowledge and Information Systems – JURIX 2021: The Thirty-fourth Annual Conference, Vilnius, Lithuania, 8-10 December 2021, 2021, p. 225–234. doi:10.3233/FAIA210341
[BibTeX] [Abstract] [Download PDF]
The justification of an algorithm’s outcomes is important in many domains, and in particular in the law. However, previous research has shown that machine learning systems can make the right decisions for the wrong reasons: despite high accuracies, not all of the conditions that define the domain of the training data are learned. In this study, we investigate what the system does learn, using state-of-the-art explainable AI techniques. With the use of SHAP and LIME, we are able to show which features impact the decision making process and how the impact changes with different distributions of the training data. However, our results also show that even high accuracy and good relevant feature detection are no guarantee for a sound rationale. Hence these state-of-the-art explainable AI techniques cannot be used to fully expose unsound rationales, further advocating the need for a separate method for rationale evaluation.
@inproceedings{DBLP:conf/jurix/StegingRV21, author = {Cor Steging and Silja Renooij and Bart Verheij}, editor = {Schweighofer Erich}, title = {Rationale Discovery and Explainable {AI}}, booktitle = {Legal Knowledge and Information Systems - {JURIX} 2021: The Thirty-fourth Annual Conference, Vilnius, Lithuania, 8-10 December 2021}, series = {Frontiers in Artificial Intelligence and Applications}, volume = {346}, pages = {225--234}, publisher = {{IOS} Press}, year = {2021}, url = {https://doi.org/10.3233/FAIA210341}, doi = {10.3233/FAIA210341}, abstract = {The justification of an algorithm’s outcomes is important in many domains, and in particular in the law. However, previous research has shown that machine learning systems can make the right decisions for the wrong reasons: despite high accuracies, not all of the conditions that define the domain of the training data are learned. In this study, we investigate what the system does learn, using state-of-the-art explainable AI techniques. With the use of SHAP and LIME, we are able to show which features impact the decision making process and how the impact changes with different distributions of the training data. However, our results also show that even high accuracy and good relevant feature detection are no guarantee for a sound rationale. Hence these state-of-the-art explainable AI techniques cannot be used to fully expose unsound rationales, further advocating the need for a separate method for rationale evaluation.} } - H. Loan, “Knowledge Representation Formalisms for Hybrid Intelligence,” in Doctoral consortium of the Knowledge Representation Conference, KR2021, online, 2021.
[BibTeX] [Abstract] [Download PDF]
Knowledge graphs can play an important role to store and provide access to global knowledge, common and accessible to both human and artificial agents, and store local knowledge of individual agents in a larger network of agents. Studying suitable formalisms to model complex, conflicting, dynamic and contextualised knowledge is still a big challenge. Therefore, we investigate the usage of knowledge representation formalisms to allow artificial intelligence systems adapt and work with complex, conflicting, dynamic and contextualized knowledge.
@inproceedings{Loan2021, address = {online}, author = {Loan, Ho}, publisher = {KR Conference}, title = {Knowledge Representation Formalisms for Hybrid Intelligence}, year = {2021}, booktitle = {Doctoral consortium of the Knowledge Representation Conference, {KR}2021}, url = "https://sites.google.com/view/kr2021dc", abstract = "Knowledge graphs can play an important role to store and provide access to global knowledge, common and accessible to both human and artificial agents, and store local knowledge of individual agents in a larger network of agents. Studying suitable formalisms to model complex, conflicting, dynamic and contextualised knowledge is still a big challenge. Therefore, we investigate the usage of knowledge representation formalisms to allow artificial intelligence systems adapt and work with complex, conflicting, dynamic and contextualized knowledge." } - B. H. Kargar, K. Miras, and A. Eiben, “The effect of selecting for different behavioral traits on the evolved gaits of modular robots,” in ALIFE 2021: The 2021 Conference on Artificial Life, 2021.
[BibTeX] [Abstract] [Download PDF]
Moving around in the environment is a fundamental skill for mobile robots. This makes the evolution of an appropriate gait, a pivotal problem in evolutionary robotics. Whereas the majority of the related studies concern robots with predefined modular or legged morphologies and locomotion speed as the optimization objective, here we investigate robots with evolvable morphologies and behavioral traits included in the fitness function. To analyze the effects we consider morphological as well as behavioral features of the evolved robots. To this end, we introduce novel behavioral measures that describe how the robot locomotes and look into the trade-off between them. Our main goal is to gain insights into differences in possible gaits of modular robots and to provide tools to steer evolution towards objectives beyond ‘simple’ speed.
@inproceedings{kargar2021effect, title = {The effect of selecting for different behavioral traits on the evolved gaits of modular robots}, author = {Kargar, Babak H and Miras, Karine and Eiben, AE}, booktitle = {ALIFE 2021: The 2021 Conference on Artificial Life}, year = {2021}, organization = {MIT Press}, url = {https://direct.mit.edu/isal/proceedings/isal/33/26/102968}, abstract = {Moving around in the environment is a fundamental skill for mobile robots. This makes the evolution of an appropriate gait, a pivotal problem in evolutionary robotics. Whereas the majority of the related studies concern robots with predefined modular or legged morphologies and locomotion speed as the optimization objective, here we investigate robots with evolvable morphologies and behavioral traits included in the fitness function. To analyze the effects we consider morphological as well as behavioral features of the evolved robots. To this end, we introduce novel behavioral measures that describe how the robot locomotes and look into the trade-off between them. Our main goal is to gain insights into differences in possible gaits of modular robots and to provide tools to steer evolution towards objectives beyond 'simple' speed.} } - G. Boomgaard, S. B. Santamaría, I. Tiddi, R. J. Sips, and Z. Szávik, “Learning profile-based recommendations for medical search auto-complete,” in AAAI-MAKE 2021 Combining Machine Learning and Knowledge Engineering, 2021, p. 1–13.
[BibTeX] [Abstract] [Download PDF]
Query popularity is a main feature in web-search auto-completion. Several personalization features have been proposed to support specific users’ searches, but often do not meet the privacy requirements of a medical environment (e.g. clinical trial search). Furthermore, in such specialized domains, the differences in user expertise and the domain-specific language users employ are far more widespread than in web-search. We propose a query auto-completion method based on different relevancy and diversity features, which can appropriately meet different user needs. Our method incorporates indirect popularity measures, along with graph topology and semantic features. An evolutionary algorithm optimizes relevance, diversity, and coverage to return a top-k list of query completions to the user. We evaluated our approach quantitatively and qualitatively using query log data from a clinical trial search engine, comparing the effects of different relevancy and diversity settings using domain experts. We found that syntax-based diversity has more impact on effectiveness and efficiency, graph-based diversity shows a more compact list of results, and relevancy the most effect on indicated preferences.
@inproceedings{boomgaard-etal-2021-learning, title = "Learning profile-based recommendations for medical search auto-complete", author = "Guusje Boomgaard and Selene Baez Santamaría and Ilaria Tiddi and Robert Jan Sips and Zoltán Szávik", keywords = "Knowledge graphs, Medical information retrieval, Professional search, Query auto-Completion", year = "2021", month = apr, day = "10", language = "English", series = "CEUR Workshop Proceedings", publisher = "CEUR-WS", pages = "1--13", editor = "Andreas Martin and Knut Hinkelmann and Hans-Georg Fill and Aurona Gerber and Doug Lenat and Reinhard Stolle and {van Harmelen}, Frank", booktitle = "AAAI-MAKE 2021 Combining Machine Learning and Knowledge Engineering", Url = "http://ceur-ws.org/Vol-2846/paper34.pdf", abstract = "Query popularity is a main feature in web-search auto-completion. Several personalization features have been proposed to support specific users' searches, but often do not meet the privacy requirements of a medical environment (e.g. clinical trial search). Furthermore, in such specialized domains, the differences in user expertise and the domain-specific language users employ are far more widespread than in web-search. We propose a query auto-completion method based on different relevancy and diversity features, which can appropriately meet different user needs. Our method incorporates indirect popularity measures, along with graph topology and semantic features. An evolutionary algorithm optimizes relevance, diversity, and coverage to return a top-k list of query completions to the user. We evaluated our approach quantitatively and qualitatively using query log data from a clinical trial search engine, comparing the effects of different relevancy and diversity settings using domain experts. We found that syntax-based diversity has more impact on effectiveness and efficiency, graph-based diversity shows a more compact list of results, and relevancy the most effect on indicated preferences.", } - S. Baez Santamaria, T. Baier, T. Kim, L. Krause, J. Kruijt, and P. Vossen, “EMISSOR: A platform for capturing multimodal interactions as Episodic Memories and Interpretations with Situated Scenario-based Ontological References,” in Proceedings of the 1st Workshop on Multimodal Semantic Representations (MMSR), Groningen, Netherlands (Online), 2021, p. 56–77.
[BibTeX] [Abstract] [Download PDF]
We present EMISSOR: a platform to capture multimodal interactions as recordings of episodic experiences with explicit referential interpretations that also yield an episodic Knowledge Graph (eKG). The platform stores streams of multiple modalities as parallel signals. Each signal is segmented and annotated independently with interpretation. Annotations are eventually mapped to explicit identities and relations in the eKG. As we ground signal segments from different modalities to the same instance representations, we also ground different modalities across each other. Unique to our eKG is that it accepts different interpretations across modalities, sources and experiences and supports reasoning over conflicting information and uncertainties that may result from multimodal experiences. EMISSOR can record and annotate experiments in virtual and real-world, combine data, evaluate system behavior and their performance for preset goals but also model the accumulation of knowledge and interpretations in the Knowledge Graph as a result of these episodic experiences.
@inproceedings{baez-santamaria-etal-2021-emissor, title = "{EMISSOR}: A platform for capturing multimodal interactions as Episodic Memories and Interpretations with Situated Scenario-based Ontological References", author = "Baez Santamaria, Selene and Baier, Thomas and Kim, Taewoon and Krause, Lea and Kruijt, Jaap and Vossen, Piek", booktitle = "Proceedings of the 1st Workshop on Multimodal Semantic Representations (MMSR)", month = jun, year = "2021", address = "Groningen, Netherlands (Online)", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2021.mmsr-1.6", pages = "56--77", abstract = "We present EMISSOR: a platform to capture multimodal interactions as recordings of episodic experiences with explicit referential interpretations that also yield an episodic Knowledge Graph (eKG). The platform stores streams of multiple modalities as parallel signals. Each signal is segmented and annotated independently with interpretation. Annotations are eventually mapped to explicit identities and relations in the eKG. As we ground signal segments from different modalities to the same instance representations, we also ground different modalities across each other. Unique to our eKG is that it accepts different interpretations across modalities, sources and experiences and supports reasoning over conflicting information and uncertainties that may result from multimodal experiences. EMISSOR can record and annotate experiments in virtual and real-world, combine data, evaluate system behavior and their performance for preset goals but also model the accumulation of knowledge and interpretations in the Knowledge Graph as a result of these episodic experiences.", } - R. Dobbe, T. Krendl Gilbert, and Y. Mintz, “Hard choices in artificial intelligence,” Artificial Intelligence, vol. 300, 2021. doi:10.1016/j.artint.2021.103555
[BibTeX] [Abstract] [Download PDF]
As AI systems are integrated into high stakes social domains, researchers now examine how to design and operate them in a safe and ethical manner. However, the criteria for identifying and diagnosing safety risks in complex social contexts remain unclear and contested. In this paper, we examine the vagueness in debates about the safety and ethical behavior of AI systems. We show how this vagueness cannot be resolved through mathematical formalism alone, instead requiring deliberation about the politics of development as well as the context of deployment. Drawing from a new sociotechnical lexicon, we redefine vagueness in terms of distinct design challenges at key stages in AI system development. The resulting framework of Hard Choices in Artificial Intelligence (HCAI) empowers developers by 1. identifying points of overlap between design decisions and major sociotechnical challenges; 2. motivating the creation of stakeholder feedback channels so that safety issues can be exhaustively addressed. As such, HCAI contributes to a timely debate about the status of AI development in democratic societies, arguing that deliberation should be the goal of AI Safety, not just the procedure by which it is ensured.
@article{dobbe_hard_2021, title = {Hard choices in artificial intelligence}, volume = {300}, issn = {0004-3702}, url = {https://www.sciencedirect.com/science/article/pii/S0004370221001065}, doi = {10.1016/j.artint.2021.103555}, abstract = {As AI systems are integrated into high stakes social domains, researchers now examine how to design and operate them in a safe and ethical manner. However, the criteria for identifying and diagnosing safety risks in complex social contexts remain unclear and contested. In this paper, we examine the vagueness in debates about the safety and ethical behavior of AI systems. We show how this vagueness cannot be resolved through mathematical formalism alone, instead requiring deliberation about the politics of development as well as the context of deployment. Drawing from a new sociotechnical lexicon, we redefine vagueness in terms of distinct design challenges at key stages in AI system development. The resulting framework of Hard Choices in Artificial Intelligence (HCAI) empowers developers by 1. identifying points of overlap between design decisions and major sociotechnical challenges; 2. motivating the creation of stakeholder feedback channels so that safety issues can be exhaustively addressed. As such, HCAI contributes to a timely debate about the status of AI development in democratic societies, arguing that deliberation should be the goal of AI Safety, not just the procedure by which it is ensured.}, language = {en}, urldate = {2021-08-04}, journal = {Artificial Intelligence}, author = {Dobbe, Roel and Krendl Gilbert, Thomas and Mintz, Yonatan}, month = nov, year = {2021}, keywords = {AI ethics, AI governance, AI regulation, AI safety, Philosophy of artificial intelligence, Sociotechnical systems} } - T. Koopman and S. Renooij, “Persuasive Contrastive Explanations for Bayesian Networks,” in Proceedings of the Sixteenth European Conference on Symbolic and Quantitative Approached to Reasoning with Uncertainty (ECSQARU), 2021, p. 229–242. doi:10.1007/978-3-030-86772-0\{_}17
[BibTeX] [Abstract] [Download PDF]
Explanation in Artificial Intelligence is often focused on providing reasons for why a model under consideration and its outcome are correct. Recently, research in explainable machine learning has initiated a shift in focus on including so-called counterfactual explanations. In this paper we propose to combine both types of explanation in the context of explaining Bayesian networks. To this end we introduce persuasive contrastive explanations that aim to provide an answer to the question Why outcome t instead of t’? posed by a user. In addition, we propose an algorithm for computing persuasive contrastive explanations. Both our definition of persuasive contrastive explanation and the proposed algorithm can be employed beyond the current scope of Bayesian networks.
@InProceedings{ecsqaru-koopman21, title = {Persuasive Contrastive Explanations for {B}ayesian Networks}, author = {Koopman, Tara AND Renooij, Silja}, booktitle = { Proceedings of the Sixteenth European Conference on Symbolic and Quantitative Approached to Reasoning with Uncertainty (ECSQARU)}, year = {2021}, pages = {229--242}, editor = {Vejnarova, J. and Wilson, N.}, volume = {12897}, series = {Lecture Notes in Computer Science}, month = {21--24 Sept}, publisher = {Springer, Cham}, pdf = {https://webspace.science.uu.nl/~renoo101/Prof/PDF/Conf/ecsqaru2021-final.pdf}, doi = {10.1007/978-3-030-86772-0\{_}17}, abstract = {Explanation in Artificial Intelligence is often focused on providing reasons for why a model under consideration and its outcome are correct. Recently, research in explainable machine learning has initiated a shift in focus on including so-called counterfactual explanations. In this paper we propose to combine both types of explanation in the context of explaining Bayesian networks. To this end we introduce persuasive contrastive explanations that aim to provide an answer to the question Why outcome t instead of t'? posed by a user. In addition, we propose an algorithm for computing persuasive contrastive explanations. Both our definition of persuasive contrastive explanation and the proposed algorithm can be employed beyond the current scope of Bayesian networks.}, URL = {https://webspace.science.uu.nl/~renoo101/Prof/PDF/Conf/ecsqaru2021-final.pdf} } - T. Koopman and S. Renooij, “Persuasive Contrastive Explanations (Extended Abstract),” in Proceedings of The 2nd Workshop in Explainable Logic-based Knowledge Representation (XLoKR), 2021.
[BibTeX] [Abstract] [Download PDF]
Explanation in Artificial Intelligence is often focused on providing reasons for why a model under consideration and its outcome are correct. Recently, research in explainable machine learning has initiated a shift in focus on including so-called counterfactual explanations. In this paper we propose to combine both types of explanation into a persuasive contrastive explanation that aims to provide an answer to the question Why outcome t instead of t’? posed by a user. In addition, we propose a model-agnostic algorithm for computing persuasive contrastive explanations from AI systems with few input variables.
@InProceedings{xlokr-koopman21, title = {Persuasive Contrastive Explanations (Extended Abstract)}, author = {Koopman, Tara AND Renooij, Silja}, booktitle = {Proceedings of The 2nd Workshop in Explainable Logic-based Knowledge Representation (XLoKR)}, year = {2021}, editor = {Baader, F. AND Bogaerts, B. AND Brewka, G. AND Hoffmann, J. AND Lukasiewicz, T. AND Potyka, N. AND Toni, F.}, month = {04--05 NOV}, pdf = {https://xlokr21.ai.vub.ac.be/papers/16/paper.pdf}, abstract = {Explanation in Artificial Intelligence is often focused on providing reasons for why a model under consideration and its outcome are correct. Recently, research in explainable machine learning has initiated a shift in focus on including so-called counterfactual explanations. In this paper we propose to combine both types of explanation into a persuasive contrastive explanation that aims to provide an answer to the question Why outcome t instead of t'? posed by a user. In addition, we propose a model-agnostic algorithm for computing persuasive contrastive explanations from AI systems with few input variables.}, URL = {https://xlokr21.ai.vub.ac.be/papers/16/paper.pdf}, } - M. A. Rahman, N. Hopner, F. Christianos, and S. V. Albrecht, “Towards Open Ad Hoc Teamwork Using Graph-based Policy Learning,” in Proceedings of the 38th International Conference on Machine Learning, 2021, p. 8776–8786.
[BibTeX] [Abstract] [Download PDF]
Ad hoc teamwork is the challenging problem of designing an autonomous agent which can adapt quickly to collaborate with teammates without prior coordination mechanisms, including joint training. Prior work in this area has focused on closed teams in which the number of agents is fixed. In this work, we consider open teams by allowing agents with different fixed policies to enter and leave the environment without prior notification. Our solution builds on graph neural networks to learn agent models and joint-action value models under varying team compositions. We contribute a novel action-value computation that integrates the agent model and joint-action value model to produce action-value estimates. We empirically demonstrate that our approach successfully models the effects other agents have on the learner, leading to policies that robustly adapt to dynamic team compositions and significantly outperform several alternative methods.
@InProceedings{pmlr-v139-rahman21a, title = {Towards Open Ad Hoc Teamwork Using Graph-based Policy Learning}, author = {Rahman, Muhammad A and Hopner, Niklas and Christianos, Filippos and Albrecht, Stefano V}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {8776--8786}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/rahman21a/rahman21a.pdf}, url = {https://proceedings.mlr.press/v139/rahman21a.html}, abstract = {Ad hoc teamwork is the challenging problem of designing an autonomous agent which can adapt quickly to collaborate with teammates without prior coordination mechanisms, including joint training. Prior work in this area has focused on closed teams in which the number of agents is fixed. In this work, we consider open teams by allowing agents with different fixed policies to enter and leave the environment without prior notification. Our solution builds on graph neural networks to learn agent models and joint-action value models under varying team compositions. We contribute a novel action-value computation that integrates the agent model and joint-action value model to produce action-value estimates. We empirically demonstrate that our approach successfully models the effects other agents have on the learner, leading to policies that robustly adapt to dynamic team compositions and significantly outperform several alternative methods.} }
2020
- B. Verheij, “Artificial intelligence as law,” Artif. Intell. Law, vol. 28, iss. 2, p. 181–206, 2020. doi:10.1007/s10506-020-09266-0
[BibTeX] [Abstract] [Download PDF]
Information technology is so ubiquitous and AI’s progress so inspiring that also legal professionals experience its benefits and have high expectations. At the same time, the powers of AI have been rising so strongly that it is no longer obvious that AI applications (whether in the law or elsewhere) help promoting a good society; in fact they are sometimes harmful. Hence many argue that safeguards are needed for AI to be trustworthy, social, responsible, humane, ethical. In short: AI should be good for us. But how to establish proper safeguards for AI? One strong answer readily available is: consider the problems and solutions studied in AI & Law. AI & Law has worked on the design of social, explainable, responsible AI aligned with human values for decades already, AI & Law addresses the hardest problems across the breadth of AI (in reasoning, knowledge, learning and language), and AI & Law inspires new solutions (argumentation, schemes and norms, rules and cases, interpretation). It is argued that the study of AI as Law supports the development of an AI that is good for us, making AI & Law more relevant than ever.
@article{Verheij20, author = {Bart Verheij}, title = {Artificial intelligence as law}, journal = {Artif. Intell. Law}, volume = {28}, number = {2}, pages = {181--206}, year = {2020}, url = {https://doi.org/10.1007/s10506-020-09266-0}, doi = {10.1007/s10506-020-09266-0}, timestamp = {Fri, 05 Jun 2020 17:08:42 +0200}, biburl = {https://dblp.org/rec/journals/ail/Verheij20.bib}, bibsource = {dblp computer science bibliography, https://dblp.org}, abstract = {Information technology is so ubiquitous and AI’s progress so inspiring that also legal professionals experience its benefits and have high expectations. At the same time, the powers of AI have been rising so strongly that it is no longer obvious that AI applications (whether in the law or elsewhere) help promoting a good society; in fact they are sometimes harmful. Hence many argue that safeguards are needed for AI to be trustworthy, social, responsible, humane, ethical. In short: AI should be good for us. But how to establish proper safeguards for AI? One strong answer readily available is: consider the problems and solutions studied in AI & Law. AI & Law has worked on the design of social, explainable, responsible AI aligned with human values for decades already, AI & Law addresses the hardest problems across the breadth of AI (in reasoning, knowledge, learning and language), and AI & Law inspires new solutions (argumentation, schemes and norms, rules and cases, interpretation). It is argued that the study of AI as Law supports the development of an AI that is good for us, making AI & Law more relevant than ever.} } - N. Kökciyan and P. Yolum, “TURP: Managing Trust for Regulating Privacy in Internet of Things,” IEEE Internet Computing, vol. 24, iss. 6, pp. 9-16, 2020. doi:https://doi.org/10.1109/MIC.2020.3020006
[BibTeX] [Abstract] [Download PDF]
Internet of Things [IoT] applications, such as smart home or ambient assisted livingsystems, promise useful services to end users. Most of these services rely heavily on sharingand aggregating information among devices; many times raising privacy concerns. Contrary totraditional systems, where privacy of each user is managed through well-defined policies, thescale, dynamism, and heterogeneity of the IoT systems make it impossible to specify privacypolicies for all possible situations. Alternatively, this paper argues that handling of privacy has tobe reasoned by the IoT devices, depending on the norms, context, as well as the trust amongentities. We present a technique, where an IoT device collects information from others, evaluatesthe trustworthiness of the information sources to decide the suitability of sharing informationwith others. We demonstrate the applicability of the technique over an IoT pilot study.
@ARTICLE{turp-ic-2020, author = {K\"okciyan, Nadin and Yolum, P{\i}nar}, journal = {IEEE Internet Computing}, title = {TURP: Managing Trust for Regulating Privacy in Internet of Things}, year = {2020}, volume = {24}, number = {6}, pages = {9-16}, abstract = {Internet of Things [IoT] applications, such as smart home or ambient assisted livingsystems, promise useful services to end users. Most of these services rely heavily on sharingand aggregating information among devices; many times raising privacy concerns. Contrary totraditional systems, where privacy of each user is managed through well-defined policies, thescale, dynamism, and heterogeneity of the IoT systems make it impossible to specify privacypolicies for all possible situations. Alternatively, this paper argues that handling of privacy has tobe reasoned by the IoT devices, depending on the norms, context, as well as the trust amongentities. We present a technique, where an IoT device collects information from others, evaluatesthe trustworthiness of the information sources to decide the suitability of sharing informationwith others. We demonstrate the applicability of the technique over an IoT pilot study.}, url = {https://webspace.science.uu.nl/~yolum001/papers/InternetComputing-20-TURP.pdf}, doi = {https://doi.org/10.1109/MIC.2020.3020006} } - O. Ulusoy and P. Yolum, “Agents for Preserving Privacy: Learning and Decision Making Collaboratively,” in Multi-Agent Systems and Agreement Technologies, 2020, p. 116–131. doi:https://doi.org/10.1007/978-3-030-66412-1_8
[BibTeX] [Abstract] [Download PDF]
Privacy is a right of individuals to keep personal information to themselves. Often online systems enable their users to select what information they would like to share with others and what information to keep private. When an information pertains only to a single individual, it is possible to preserve privacy by providing the right access options to the user. However, when an information pertains to multiple individuals, such as a picture of a group of friends or a collaboratively edited document, deciding how to share this information and with whom is challenging as individuals might have conflicting privacy constraints. Resolving this problem requires an automated mechanism that takes into account the relevant individuals’ concerns to decide on the privacy configuration of information. Accordingly, this paper proposes an auction-based privacy mechanism to manage the privacy of users when information related to multiple individuals are at stake. We propose to have a software agent that acts on behalf of each user to enter privacy auctions, learn the subjective privacy valuations of the individuals over time, and to bid to respect their privacy. We show the workings of our proposed approach over multiagent simulations.
@InProceedings{ulusoy-yolum-20, title = "Agents for Preserving Privacy: Learning and Decision Making Collaboratively", author = "Ulusoy, Onuralp and Yolum, P{\i}nar", editor = "Bassiliades, Nick and Chalkiadakis, Georgios and de Jonge, Dave", booktitle = "Multi-Agent Systems and Agreement Technologies", year = "2020", publisher = "Springer International Publishing", pages = "116--131", abstract = "Privacy is a right of individuals to keep personal information to themselves. Often online systems enable their users to select what information they would like to share with others and what information to keep private. When an information pertains only to a single individual, it is possible to preserve privacy by providing the right access options to the user. However, when an information pertains to multiple individuals, such as a picture of a group of friends or a collaboratively edited document, deciding how to share this information and with whom is challenging as individuals might have conflicting privacy constraints. Resolving this problem requires an automated mechanism that takes into account the relevant individuals' concerns to decide on the privacy configuration of information. Accordingly, this paper proposes an auction-based privacy mechanism to manage the privacy of users when information related to multiple individuals are at stake. We propose to have a software agent that acts on behalf of each user to enter privacy auctions, learn the subjective privacy valuations of the individuals over time, and to bid to respect their privacy. We show the workings of our proposed approach over multiagent simulations.", isbn = "978-3-030-66412-1", doi = {https://doi.org/10.1007/978-3-030-66412-1_8}, url = {https://webspace.science.uu.nl/~yolum001/papers/ulusoy-yolum-20.pdf} } - L. Krause and P. Vossen, “When to explain: Identifying explanation triggers in human-agent interaction,” in 2nd Workshop on Interactive Natural Language Technology for Explainable Artificial Intelligence, Dublin, Ireland, 2020, p. 55–60.
[BibTeX] [Abstract] [Download PDF]
With more agents deployed than ever, users need to be able to interact and cooperate with them in an effective and comfortable manner. Explanations have been shown to increase the understanding and trust of a user in human-agent interaction. There have been numerous studies investigating this effect, but they rely on the user explicitly requesting an explanation. We propose a first overview of when an explanation should be triggered and show that there are many instances that would be missed if the agent solely relies on direct questions. For this, we differentiate between direct triggers such as commands or questions and introduce indirect triggers like confusion or uncertainty detection.
@inproceedings{krause-vossen-2020-explain, title = "When to explain: Identifying explanation triggers in human-agent interaction", author = "Krause, Lea and Vossen, Piek", booktitle = "2nd Workshop on Interactive Natural Language Technology for Explainable Artificial Intelligence", month = nov, year = "2020", address = "Dublin, Ireland", publisher = "Association for Computational Linguistics", url = "https://www.aclweb.org/anthology/2020.nl4xai-1.12", pages = "55--60", abstract = "With more agents deployed than ever, users need to be able to interact and cooperate with them in an effective and comfortable manner. Explanations have been shown to increase the understanding and trust of a user in human-agent interaction. There have been numerous studies investigating this effect, but they rely on the user explicitly requesting an explanation. We propose a first overview of when an explanation should be triggered and show that there are many instances that would be missed if the agent solely relies on direct questions. For this, we differentiate between direct triggers such as commands or questions and introduce indirect triggers like confusion or uncertainty detection.", } - P. K. Murukannaiah, N. Ajmeri, C. M. Jonker, and M. P. Singh, “New Foundations of Ethical Multiagent Systems,” in Proceedings of the 19th Conference on Autonomous Agents and MultiAgent Systems, Auckland, 2020, p. 1706–1710.
[BibTeX] [Abstract] [Download PDF]
Ethics is inherently a multiagent concern. However, research on AI ethics today is dominated by work on individual agents: (1) how an autonomous robot or car may harm or (differentially) benefit people in hypothetical situations (the so-called trolley problems) and (2) how a machine learning algorithm may produce biased decisions or recommendations. The societal framework is largely omitted. To develop new foundations for ethics in AI, we adopt a sociotechnical stance in which agents (as technical entities) help autonomous social entities or principals (people and organizations). This multiagent conception of a sociotechnical system (STS) captures how ethical concerns arise in the mutual interactions of multiple stakeholders. These foundations would enable us to realize ethical STSs that incorporate social and technical controls to respect stated ethical postures of the agents in the STSs. The envisioned foundations require new thinking, along two broad themes, on how to realize (1) an STS that reflects its stakeholders’ values and (2) individual agents that function effectively in such an STS.
@inproceedings{Murukannaiah-2020-AAMASBlueSky-EthicalMAS, author = {Pradeep K. Murukannaiah and Nirav Ajmeri and Catholijn M. Jonker and Munindar P. Singh}, title = {New Foundations of Ethical Multiagent Systems}, booktitle = {Proceedings of the 19th Conference on Autonomous Agents and MultiAgent Systems}, series = {AAMAS '20}, year = {2020}, address = {Auckland}, pages = {1706--1710}, numpages = {5}, keywords = {Agents, ethics}, url = {https://ii.tudelft.nl/~pradeep/doc/Murukannaiah-2020-AAMASBlueSky-EthicalMAS.pdf}, abstract = {Ethics is inherently a multiagent concern. However, research on AI ethics today is dominated by work on individual agents: (1) how an autonomous robot or car may harm or (differentially) benefit people in hypothetical situations (the so-called trolley problems) and (2) how a machine learning algorithm may produce biased decisions or recommendations. The societal framework is largely omitted. To develop new foundations for ethics in AI, we adopt a sociotechnical stance in which agents (as technical entities) help autonomous social entities or principals (people and organizations). This multiagent conception of a sociotechnical system (STS) captures how ethical concerns arise in the mutual interactions of multiple stakeholders. These foundations would enable us to realize ethical STSs that incorporate social and technical controls to respect stated ethical postures of the agents in the STSs. The envisioned foundations require new thinking, along two broad themes, on how to realize (1) an STS that reflects its stakeholders' values and (2) individual agents that function effectively in such an STS.} } - Z. Akata, D. Balliet, M. de Rijke, F. Dignum, V. Dignum, G. Eiben, A. Fokkens, D. Grossi, K. Hindriks, H. Hoos, H. Hung, C. Jonker, C. Monz, M. Neerincx, F. Oliehoek, H. Prakken, S. Schlobach, L. van der Gaag, F. van Harmelen, H. van Hoof, B. van Riemsdijk, A. van Wynsberghe, R. Verbrugge, B. Verheij, P. Vossen, and M. Welling, “A Research Agenda for Hybrid Intelligence: Augmenting Human Intellect With Collaborative, Adaptive, Responsible, and Explainable Artificial Intelligence,” IEEE Computer, vol. 53, iss. 08, pp. 18-28, 2020. doi:10.1109/MC.2020.2996587
[BibTeX] [Abstract] [Download PDF]
We define hybrid intelligence (HI) as the combination of human and machine intelligence, augmenting human intellect and capabilities instead of replacing them and achieving goals that were unreachable by either humans or machines. HI is an important new research focus for artificial intelligence, and we set a research agenda for HI by formulating four challenges.
@ARTICLE {9153877, author = {Z. Akata and D. Balliet and M. de Rijke and F. Dignum and V. Dignum and G. Eiben and A. Fokkens and D. Grossi and K. Hindriks and H. Hoos and H. Hung and C. Jonker and C. Monz and M. Neerincx and F. Oliehoek and H. Prakken and S. Schlobach and L. van der Gaag and F. van Harmelen and H. van Hoof and B. van Riemsdijk and A. van Wynsberghe and R. Verbrugge and B. Verheij and P. Vossen and M. Welling}, journal = {IEEE Computer}, title = {A Research Agenda for Hybrid Intelligence: Augmenting Human Intellect With Collaborative, Adaptive, Responsible, and Explainable Artificial Intelligence}, year = {2020}, volume = {53}, number = {08}, issn = {1558-0814}, pages = {18-28}, doi = {10.1109/MC.2020.2996587}, publisher = {IEEE Computer Society}, address = {Los Alamitos, CA, USA}, month = {aug}, url = "http://www.cs.vu.nl/~frankh/postscript/IEEEComputer2020.pdf", abstract = "We define hybrid intelligence (HI) as the combination of human and machine intelligence, augmenting human intellect and capabilities instead of replacing them and achieving goals that were unreachable by either humans or machines. HI is an important new research focus for artificial intelligence, and we set a research agenda for HI by formulating four challenges." } - B. M. Renting, H. H. Hoos, and C. M. Jonker, “Automated Configuration of Negotiation Strategies,” in Proceedings of the 19th International Conference on Autonomous Agents and MultiAgent Systems, 2020, p. 1116–1124.
[BibTeX] [Abstract] [Download PDF]
Bidding and acceptance strategies have a substantial impact on the outcome of negotiations in scenarios with linear additive and nonlinear utility functions. Over the years, it has become clear that there is no single best strategy for all negotiation settings, yet many fixed strategies are still being developed. We envision a shift in the strategy design question from: What is a good strategy?, towards: What could be a good strategy? For this purpose, we developed a method leveraging automated algorithm configuration to find the best strategies for a specific set of negotiation settings. By empowering automated negotiating agents using automated algorithm configuration, we obtain a flexible negotiation agent that can be configured automatically for a rich space of opponents and negotiation scenarios. To critically assess our approach, the agent was tested in an ANAC-like bilateral automated negotiation tournament setting against past competitors. We show that our automatically configured agent outperforms all other agents, with a 5.1 percent increase in negotiation payoff compared to the next-best agent. We note that without our agent in the tournament, the top-ranked agent wins by a margin of only 0.01 percent .
@inproceedings{Renting2020AutomatedStrategies, title = {Automated Configuration of Negotiation Strategies}, booktitle = {Proceedings of the 19th International Conference on Autonomous Agents and MultiAgent Systems}, author = {Renting, Bram M. and Hoos, Holger H. and Jonker, Catholijn M.}, year = {2020}, month = {may}, series = {AAMAS '20}, pages = {1116--1124}, publisher = {International Foundation for Autonomous Agents and Multiagent Systems}, abstract = {Bidding and acceptance strategies have a substantial impact on the outcome of negotiations in scenarios with linear additive and nonlinear utility functions. Over the years, it has become clear that there is no single best strategy for all negotiation settings, yet many fixed strategies are still being developed. We envision a shift in the strategy design question from: What is a good strategy?, towards: What could be a good strategy? For this purpose, we developed a method leveraging automated algorithm configuration to find the best strategies for a specific set of negotiation settings. By empowering automated negotiating agents using automated algorithm configuration, we obtain a flexible negotiation agent that can be configured automatically for a rich space of opponents and negotiation scenarios. To critically assess our approach, the agent was tested in an ANAC-like bilateral automated negotiation tournament setting against past competitors. We show that our automatically configured agent outperforms all other agents, with a 5.1 percent increase in negotiation payoff compared to the next-best agent. We note that without our agent in the tournament, the top-ranked agent wins by a margin of only 0.01 percent .}, isbn = {978-1-4503-7518-4}, keywords = {automated algorithm configuration,automated negotiation,negotiation strategy}, url = {https://ifaamas.org/Proceedings/aamas2020/pdfs/p1116.pdf} }
- B. M. Renting, H. H. Hoos, and C. M. Jonker, “Automated Configuration and Usage of Strategy Portfolios for Bargaining.” 2021-12-14. doi:10.48550/arXiv.2212.10228
[BibTeX] [Abstract] [Download PDF]
Bargaining can be used to resolve mixed-motive games in multi-agent systems. Although there is an abundance of negotiation strategies implemented in automated negotiating agents, most agents are based on single fixed strategies, while it is widely acknowledged that there is no single best-performing strategy for all negotiation settings. In this paper, we focus on bargaining settings where opponents are repeatedly encountered, but the bargaining problems change. We introduce a novel method that automatically creates and deploys a portfolio of complementary negotiation strategies using a training set and optimise pay-off in never-before-seen bargaining settings through per-setting strategy selection. Our method relies on the following contributions. We introduce a feature representation that captures characteristics for both the opponent and the bargaining problem. We model the behaviour of an opponent during a negotiation based on its actions, which is indicative of its negotiation strategy, in order to be more effective in future encounters. Our combination of feature-based methods generalises to new negotiation settings, as in practice, over time, it selects effective counter strategies in future encounters. Our approach is tested in an ANAC-like tournament, and we show that we are capable of winning such a tournament with a 5.6 percent increase in pay-off compared to the runner-up agent.
@inproceedings{rentingAutomatedConfigurationUsage2021, title = {Automated Configuration and Usage of Strategy Portfolios for Bargaining}, author = {Renting, Bram M. and Hoos, Holger H. and Jonker, Catholijn M.}, date = {2021-12-14}, eprint = {2212.10228}, eprinttype = {arxiv}, primaryclass = {cs}, publisher = {{arXiv}}, doi = {10.48550/arXiv.2212.10228}, url = {http://arxiv.org/abs/2212.10228}, urldate = {2022-12-29}, abstract = {Bargaining can be used to resolve mixed-motive games in multi-agent systems. Although there is an abundance of negotiation strategies implemented in automated negotiating agents, most agents are based on single fixed strategies, while it is widely acknowledged that there is no single best-performing strategy for all negotiation settings. In this paper, we focus on bargaining settings where opponents are repeatedly encountered, but the bargaining problems change. We introduce a novel method that automatically creates and deploys a portfolio of complementary negotiation strategies using a training set and optimise pay-off in never-before-seen bargaining settings through per-setting strategy selection. Our method relies on the following contributions. We introduce a feature representation that captures characteristics for both the opponent and the bargaining problem. We model the behaviour of an opponent during a negotiation based on its actions, which is indicative of its negotiation strategy, in order to be more effective in future encounters. Our combination of feature-based methods generalises to new negotiation settings, as in practice, over time, it selects effective counter strategies in future encounters. Our approach is tested in an ANAC-like tournament, and we show that we are capable of winning such a tournament with a 5.6 percent increase in pay-off compared to the runner-up agent.}, archiveprefix = {arXiv}, eventtitle = {The {{Second Cooperative AI}} Workshop, {{NeurIPS}} 2021}, keywords = {Computer Science - Multiagent Systems}, } - B. M. Renting, H. H. Hoos, and C. M. Jonker, “Automated Configuration and Usage of Strategy Portfolios for Mixed-Motive Bargaining,” in Proceedings of the 21st International Conference on Autonomous Agents and Multiagent Systems, 2022-05-09, p. 1101–1109.
[BibTeX] [Abstract] [Download PDF]
Bargaining can be used to resolve mixed-motive games in multi-agent systems. Although there is an abundance of negotiation strategies implemented in automated negotiating agents, most agents are based on single fixed strategies, while it is acknowledged that there is no single best-performing strategy for all negotiation settings. In this paper, we focus on bargaining settings where opponents are repeatedly encountered, but the bargaining problems change. We introduce a novel method that automatically creates and deploys a portfolio of complementary negotiation strategies using a training set and optimise pay-off in never-before-seen bargaining settings through per-setting strategy selection. Our method relies on the following contributions. We introduce a feature representation that captures characteristics for both the opponent and the bargaining problem. We model the behaviour of an opponent during a negotiation based on its actions, which is indicative of its negotiation strategy, in order to be more effective in future encounters. Our combination of feature-based methods generalises to new negotiation settings, as in practice, over time, it selects effective counter strategies in future encounters. Our approach is tested in an ANAC-like tournament, and we show that we are capable of winning such a tournament with a 5.6 percent increase in pay-off compared to the runner-up agent.
@inproceedings{rentingAutomatedConfigurationUsage2022, title = {Automated Configuration and Usage of Strategy Portfolios for Mixed-Motive Bargaining}, booktitle = {Proceedings of the 21st {{International Conference}} on {{Autonomous Agents}} and {{Multiagent Systems}}}, author = {Renting, Bram M. and Hoos, Holger H. and Jonker, Catholijn M.}, date = {2022-05-09}, series = {{{AAMAS}} '22}, pages = {1101--1109}, publisher = {{International Foundation for Autonomous Agents and Multiagent Systems}}, location = {{Richland, SC}}, url = {https://ifaamas.org/Proceedings/aamas2022/pdfs/p1101.pdf}, urldate = {2022-10-23}, abstract = {Bargaining can be used to resolve mixed-motive games in multi-agent systems. Although there is an abundance of negotiation strategies implemented in automated negotiating agents, most agents are based on single fixed strategies, while it is acknowledged that there is no single best-performing strategy for all negotiation settings. In this paper, we focus on bargaining settings where opponents are repeatedly encountered, but the bargaining problems change. We introduce a novel method that automatically creates and deploys a portfolio of complementary negotiation strategies using a training set and optimise pay-off in never-before-seen bargaining settings through per-setting strategy selection. Our method relies on the following contributions. We introduce a feature representation that captures characteristics for both the opponent and the bargaining problem. We model the behaviour of an opponent during a negotiation based on its actions, which is indicative of its negotiation strategy, in order to be more effective in future encounters. Our combination of feature-based methods generalises to new negotiation settings, as in practice, over time, it selects effective counter strategies in future encounters. Our approach is tested in an ANAC-like tournament, and we show that we are capable of winning such a tournament with a 5.6 percent increase in pay-off compared to the runner-up agent.}, isbn = {978-1-4503-9213-6}, keywords = {algorithm configuration,algorithm selection,bargaining,mixed-motive games}, }