Best paper award for Ilaria, Victor and Stefan
The paper “Knowledge Engineering for Hybrid Intelligence” by Ilaria Tiddi, Victor de Boer, Stefan Schlobach (all VU Amsterdam) and André Meyer-Vitali (DFKI) received the best research paper award at KCAP 2023, the conference on Knowledge Capture, one of the two leading venues worldwide on Knowledge Engineering.
The paper investigates how classical Knowledge Engineering methods can be adapted for use in HI scenarios, and proposes a new ontology for HI knowledge roles, a set of HI tasks, and a repository for HI scenario’s, allowing reuse, validation and design of existing and new HI applications.
Best paper award for Ilaria, Victor and Stefan
The paper “Knowledge Engineering for Hybrid Intelligence” by Ilaria Tiddi, Victor de Boer, Stefan Schlobach (all VU Amsterdam) and André Meyer-Vitali (DFKI) received the best research paper award at KCAP 2023, the conference on Knowledge Capture, one of the two leading venues worldwide on Knowledge Engineering.
The paper investigates how classical Knowledge Engineering methods can be adapted for use in HI scenarios, and proposes a new ontology for HI knowledge roles, a set of HI tasks, and a repository for HI scenario’s, allowing reuse, validation and design of existing and new HI applications.
Catholijn Jonker elected as NAE fellow
Catholijn Jonker was elected as one of founding Fellows of the newly formed Netherlands Academy of Engineering. The NAE unites the top experts active in technological sciences and applied research and development. Only 62 leading experts nationwide were elected as founding Fellows across all disciplines.
Booklet on AI by Bennie Mols
Science journalist Bennie Mols published a booklet for the general public on how AI can support humans instead of replacing them, with an entire chapter on Hybrid Intelligence.
Tenure for Shihan Wang
Shihan Wang got tenure at the department of Information and Computing Science, Faculty of Science, Utrecht University. She will continue her research on reinforcement learning, with a focus on learning through efficient and reliable communication. It covers the interactions between human and agent, as well as among multiple agents.
Honorable Mention at ACII 2023 for paper by HI members
The paper “Fine-grained Affective Processing Capabilities Emerging from Large Language Models” by Joost Broekens, Bernhard Hilpert, Suzan Verberne, Kim Baraka, Patrick Gebhard and Aske Plaat has won the Honorable Mention at the International Conference on Affective Computing and Intelligent Interactions (ACII 2023). Their paper is available here.
Ammodo interviews Frank van Harmelen
Frank van Harmelen was interviewed by Ammodo, an organisation that stimulates the development of arts, architecture and sciences. The interview (in dutch) discusses how human and machine can collaborate successfully.
Bernd Dudzik, Mike Ligthart, Erman Acar, and Karine Miras promoted to Assistant professors
Four of our HI Post-docs were promoted to assistant professors. Luckily, they will stay affiliated with the Hyrbid Intelligence Centre having projects to supervise in the 1st or 2nd recruitment round. We are looking forward to their much-valued contributions to the programme.
VIDI-grant for Andrew Yates
Andrew Yates won an NWO VIDI grant for his research proposal entitled ‘IDEAS: Incremental Dense Representations’. The research concerns improvements to content-based search and recommendation systems. These usually take a piece of text as an input, such as a web page or a job posting, and create a representation of this entire text that is used to return results for the user’s query or recommendation need. This approach works well, but it lacks control over how parts of the input text influence the resulting representation. Andrew proposes a paradigm where information is incrementally added into a representation piece-by-piece. This control can improve result quality and enable analysis of the content inside a representation, allowing us to look for harmful content like biases.
FoMo lecture by Selene Báez Santamaría
Selene Báez Santamaría gave a lecture in the FoMo series, which is a series of seminars highlighting the work being done in Amsterdam around foundation models: large, highly re-usable machine learning models trained on great amounts of data.
Selenes FoMo lecture is entitled ‘A Practical Approach to GPTx Models: Evaluating Task-Specific Performance and Insights’, concerning task-specific performance of large language models. Within research, the API accessibility of such models has made it possible to effortlessly explore these model’s performance on a wide range of scientific tasks. In this talk Selene presented her findings from three specific tasks: knowledge base completion, argument mining, and task-oriented dialogue incorporating subjective knowledge.
Mike Ligthart presents at TEDxVUAmsterdam
Mike Ligthart was one of the speakers of the third event of TEDxVUAmsterdam, which focussed on the theme ‘Shaping Tomorrow’. His TEDtalk, entitled ‘A Robot With a Story’, introduced the audience to Hero and Memo, two social robots. Hero wants to become champion of the robo-olympics. Memo, on the other hand, has no nose and just wants to know how the forest smells like. Through their stories, Hero is able to distract children with cancer from stressful moments in the hospital and Memo is able to elicit stories, thought forgotten, from people with dementia. These stories were created by a close collaboration between creative writers, the people interacting with the robots, and me, an artificial intelligence (AI) researcher. In this TEDtalk, Mike reflected on this collaboration, and professed his love for the profession of writers and all that it brings. By understanding, no feeling, the human experience writers are pivotal for designing meaningful interactions between people and social robots. Something AI cannot solve on its own.
The TED talk can be viewed here.
Robosurgery case study for HI team
A team of the HI consortium will have an exciting robosurgery case study meeting on June 19 at the Erasmus MC. The program includes demos, short presentations, and hands-on activities, and the team gets the chance to wear a full medical suit and try to sew a dummy blood vessel. This helps the team concretize research themes, research questions, methods, and evaluation metrics for the coming years, as well as identifying synergies with current and upcoming projects.
HI Consortium meeting 8-10 May 2023
This May, next to several different weather types, the Vlieland meeting has offered us a programme centered around co-activities. Matt Johnson presented his keynote lecture on “Co-activity analysis” and guided the subsequent exercise to try and make a co-activity analysis for our two Case Studies on Diabetes and Robotic Surgery. We enjoyed each others company while having dinner, going for beautiful walks, including the HI outdoor assignment organised by Davide Dell’Anna, bowling, swimming and watching the movies selected by Max van Duijn. We thank all Matrix Leads for organising their meetings and giving an introduction to the new structure in HI.
Pinar Yolum appointed professor of Trustworthy AI (UU)
Pinar Yolum has been appointed professor of Trustworthy AI at Utrecht University, Faculty of Science, Department of Information and Computing Sciences.
Roel Dobbe presents at UC Berkely and UC San Diego
From end of March to early May, Roel Dobbe was on a mini-sabbatical in California, spending time with this former colleagues from his time as PhD at UC Berkely. In addition, he visited UC San Diego. At both UCs he presented on his research vision and main results, entitled: ‘Algorithms as Sociotechnical Systems: A systems and control paradigm for research, engineering and governance’.
Christof Monz appointed professor of Language Technology (UvA)
Christof Monz has been appointed professor of Language Technology at the Faculty of Science of the University of Amsterdam (UvA). His professorial inauguration is on June 1st 2023, entitled “From patterns to meaning and back”.
Roel Dobbe tenured at Department of Engineering Systems and Services (TUD)
Roel Dobbe got tenure at the Faculty of Technology, Policy and Management in the Department of Engineering Systems and Services. His lab, called “Socio-Technical-Ecological perspectives on Algorithmic and Data-driven (STEAD) Systems Lab”, aims to build bridges between the technical disciplines typically involved in the foundational advances in these systems as well as their applications, and the many non-technical disciplines and actors that are or should be engaged in better understanding their impacts and design for safer and more responsible configurations.
Rineke Verbrugge chairs TARK 2023
This year, Rineke Verbrugge is the programme chair of “The Nineteenth Conference on Theoretical Aspects of Rationality and Knowledge”, TARK 2023, which will be held in Oxford, UK, end of June 2023.
Rineke Verbrugge invited for prestigious Lindström Lectures in Sweden
Rineke Verbrugge was invited to give the prestigious 2023 Lindström Lectures. The lecture, entitled ‘Aspects of provability and interpretability’ is held at University of Gothenborg, Sweden in May 2023.
HI part of winning team in the AI for Global Climate Cooperation Competition
Erman Acar, Robert Loftin and Bram Renting were part of the winning team in the AI for Global Climate Cooperation Competition. This competition focused at designing climate agreements and negotiation protocols that lead to a sustainable future.
The winning team had the highest score in the area of climate and global economy, and excelled on the real-world applicability of their developed negotiation protocols, by making the relation with international agreements of the World Trade Organization (WTO) and the Carbon Border Adjustment Mechanism (CBAM) if the European Union.
Michiel van der Meer presents at NEN on language models
On March 22, 2023, Michiel van der Meer gave a presentation for the Management Team (MT) of the Dutch Normalisation Institute (Nederlands Normalisatie Instituut, NEN) on languagemodels such as LLMs and ChatGPT. In this MT meeting also a SWOT analysis was made, in which Hybrid Intelligence was identified as an opportunity.
Henry Prakken wins Stanford CodeX 2023 prize
At CodeX FutureLaw 2023 Conference, it was announced that Henry Prakken was awarded with the 2023 CodeX Prize, an annual award given to an individual or group for a noteworthy contribution to computational law. He is one of the three winning academic researchers from three different countries who have collaborated for decades on legal argumentation theory. Their areas of research broadly impact on how artificial intelligence can be leveraged in the formulation of sophisticated legal arguments.
Siddharth Mehrotra wins book from Prof. Ben Shneiderman
With his report entitled ‘Designing for Appropriate Trust in Human-AI Interaction’, Siddharth Mehrotra was one of the winners of best report on Human-Centered AI on the topic of Designing for Appropriate Trust in Human-AI Interaction organized by Prof. Ben Shneiderman in cooperation with Oxford University Press 100th NOTE of Human-Centered AI Google Group. He will receive a copy of the book Human-Centered AI from Oxford University Press.
Siddarth Mehrotra gives keynote at ‘Authenticity in the Digital Age’ symposium in Tilburg
In April 2023, Siddarth Mehrotra gave a keynote entitled ‘Authenticity in the Digital Age: Designing for Appropriate Trust ‘at the ‘Authenticity in the Digital Age’ symposium in Tilburg. He explained how authentic data can contribute to our trust in computer systems.
Workshop on Interdisciplinary Perspectives on Technologies for Mingling
On 2 May, Hayley Hung aims to present an overarching view of the research results of the NWO funded Vidi project MINGLE (Modelling Social Group Dynamics and Interaction Quality in Complex Scenes using Multi-Sensor Analysis of Non-Verbal Behaviour). The research results of MINGLE will feed into a new ERC Consolidator grant funded project NEON (Nonverbal Intention Modelling) which will focus on the analysis of intention, particularly in mingling settings.
To participate, register here
Hayley Hung wins ERC consolidator grant
Hayley Hung has been awarded an ERC consolidator grant. The NEON project investigates how we can develop artificially intelligent systems to infer social intentions from non-verbal behaviour.
The project addresses intention detection in open-ended contexts involving large unstructured social gatherings such as networking or mingling events. During these events, there are no prearranged conversations, multiple conversations can occur at the same time, and all conversation come about via coordination with multiple independent actors with their own possibly conflicting goals. How do we train machines to perceive plausible intentions in these complex, unstructured social environments?
Frank van Harmelen gives keynote at SWAT4HCLS in Basel
Frank gave the closing keynote at the annual “Semantic Web Applications and Tools for Healthcare and Life Sciences” in Basel, on systems that combine learning and reasoning, including HI relevant architectures for using knowledge to constrain, accelerate and explain learning.
Birna van Riemsdijk among the 5 nominees for Women in AI NL
We’re proud that Birna van Riemsdijk is one of the 5 nominees for the “Women researchers in AI NL” awards. The award is dedicated to women conducting research that has an impact on the future of AI, women willing to contribute with their expertise and personality to lead the future of AI. Many congratulations to Birna for this well deserved nomination!
2nd Workshop on Human-Interactive Robot Learning
Kim Baraka is co-organising the 2nd Workshop on Human-Interactive Robot Learning, at the Human-Robot Interaction conference in Stockholm in May 2023.
Newsletter
Does AI need a body? Debate at the VU, 9 Dec
Most applications of AI do not have a face or a body and it is mostly a “hidden” system that is difficult to relate to for people. In this debate, we discuss whether AI needs a body. Are there any benefits, what are the differences with virtual AI, how difficult is it and how can we make progress in science and technology development? Several panellists will present their view and have a moderated debate with students and colleagues.
Friday 9 Dec, 15h-17h30, Aurora, HG-0C29, VU Amsterdam. See here for details
Henry Prakken invited speaker at European Anti-Fraud Office
Henri Prakken was invited speaker at the Conference on Judgment- and Decision Making in fraud and corruption investigations, organised by the European Anti-Fraud Office Brussels (Belgium), 7-8 December 2022 with a talk entitled ‘Three approaches to rational proof in criminal cases’.
HI Consortium meeting 6 February 2023
The programme of our consortium meeting in Amersfoort was centered around the Debate on the relevance of Large Language Models to HI. Each panel member, Piek Vossen, Maarten de Rijke, and Bart Verheij, presented two propositions to start a lively panel discussion with the audience.
Masha Tsfasman under the guidance of Enrico and Catholijn from the TU Delft team won the ChatGPT challenge, with “The promising use-case of ChatGPT: Analysing group deliberations and debates”. To make sure that the new projects and their initiators of our second recruitment round are well known to all consortium members we organised speed-dates to discuss the project goals and how they connect with the existing projects and initiatives.
Birna van Riemsdijk wraps up VIDI project with symposium on “Being human in the digital society: on technology, norms and us”.
Birna van Riemsdijk (Univ. of Twente) wrapped up her VIDI project with a symposium on “Being human in the digital society: on technology, norms and us”. Various HI members participated and contributed. There’s a brief report on the symposium, including a 10min video on the project.
Selene Baez member of winning team for ISWC knowledge graph construction challenge
Selene Baez from the CTLT team at the VU was member of the all-VU team that won the challenge at the International Semantic Web Conference (ISWC) on constructing symbolic knowledge graphs from neural large language models. Their paper “Prompting as probing” describing their results can be found here.
HI team wins argument mining challenge at COLING 2022
A team consisting of HI members Michiel van der Meer, Urja Khurana, Lea Krause and Selene Baez, together with their VU colleague Myrthe Reuver won the COLING 2022 challenge on predicting the validity and novelty of arguments stated in natural language as part of the 9th Workshop on Argument Mining. Their work is reported here.
Best paper at COLING 2022 workshop
Selene Baez, Piek Vossen and their colleague Thomas Baier won the best paper award at the 1st Workshop on Customized Chat Grounding Persona and Knowledge for their paper Evaluating Agent Interactions Through Episodic Knowledge Graphs
Henry Prakken gives keynote at LNGAI 2022
Henri Prakken was keynote speaker, at the Second International Workshop on Logics for New-Generation Artificial Intelligence (LNGAI 2022), Zhuhai Campus of Sun Yat-sen University, China. (online because of covid-19), speaking on ‘Principle-based and principled formal studies of argumentation: the case of gradual semantics.’
HHAI 2022 proceedings available
The proceedings of the 1st International Conference on Hybrid Human-Artificial Intelligence (HHAI 2022) are now available under open access. Download them here.
PhD defense of Mike Ligthart
“A sense of connection is essential when deploying robots for children”.
Mike Lighthart defends his PhD thesis on 30 September. See here for an explanation of his work (in Dutch)
AI learns when to leave a decision to an expert
Part of being smart is knowing where your limitations lie. Learning AI systems have long been bad at this. But thanks to UvA researchers Rajeev Verma and Eric Nalisnick, AI systems are now starting to learn when it would be better to leave a decision to a human expert. See here for a brief explanation.
Catholijn Jonker keynote speaker at IC3K 2022
Catholijn Jonker is keynote speaker at the combined IC3K conferences in Malta, on the topic “Self-Reflective Hybrid Intelligence: Combining Human with Artificial Intelligence and Logic“
HHAI 2022 Conference
Here is a video impression of our very successful First International Conference on Hybrid Human-Artificial Intelligence.
250 participants, 5 days, 7 workshops, 45 main conference papers
NIAS Fellowship for Davide Grossi
Davide has been selected for a 5 months fellowship at NIAS, the Netherlands Institute for Advanced Study in the Humanities and Social Science of the Royal Dutch Academy of Science, studying how to develop digital democracy algorithms that we can trust to be truly democratic.
Best Paper at AAMAS 2022
Davide Grossi won a best paper award at the AAMAS (International Conference on Autonomous Agents and Multiagent Systems).
The paper called “Social Choice Around the Block: On the Computational Social Choice of Blockchain” charts a research agenda at the interface of multi-agent systems and blockchain research. Davide Grossi won the award in the “Blue Sky Ideas track”, which emphasises visionary ideas, long-term challenges and new research opportunities.
Best Paper at HRI 2022
Mike Ligthart, Koen Hindriks and Mark Neerincx received the Best paper award at HRI 2022, the ACM/IEEE International Conference on Human-Robot Interaction 2022 for their paper: Memory-Based Personalization for Fostering a Long-Term Child-Robot Relationship. Impressive!
NWO Grant for Eric Nalisnick
Eric Nalisnick has won a prestigious VENI grant from NWO on “Continual Learning under Human Guidance”.
Winning the NeurIPS IGLU Challenge
Our PhD students Putra Manggala, Kata Naszadi, Michiel van der Meer and Tawewoon Kim won the IGLU contest on ‘ Interactive Grounded Language Understanding in a Collaborative Environment’ at NeurIPS 2021.
Paper of the year award
Our paper “A Research Agenda for Hybrid Intelligence” in IEEE Computer received the paper of the year award!
HHAI 2022
We are organising the 1st International Conference on Hybrid Human-Artificial Intelligence, 13-17 June 2022 in Amsterdam. See https://www.hhai-conference.org/ for more information.
