Keynote Speakers
KEYNOTE SPEAKERS LIST
Cristiano Castelfranchi, Institute of Cognitive Sciences and Technologies (ISTC), Italy
Title: The Fundamental Ontology and Dynamics of Goals in Mind and Society
Boi Faltings, Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland
Title: Getting Agents to Tell the Truth
Didier Dubois, Institut de Recherche en Informatique de Toulouse (IRIT), France
Title: Uncertainty Theories, Degrees of Truth and Epistemic States
Mark Klein, MIT Center for Collective Intelligence, United States of America
Title: The MIT Deliberatorium: Enabling Large-Scale Deliberation about Complex Systemic Problems
Klaus Fischer, Agents and Simulated Reality, DFKI GmbH, Germany
Title: Model Driven Design of Agents and Multiagent Systems
The Fundamental Ontology and Dynamics of Goals in Mind and Society
|
Cristiano Castelfranchi Institute of Cognitive Sciences and Technologies (ISTC) Italy |
Brief Bio
Cristiano Castelfranchi is Full professor of "Cognitive Sciences" at the University of Siena, Department of Communication Science, and Director of the Institute of Cognitive Sciences and Technologies- ISTC of the National Research Council, in Roma.
A cognitive scientist with a background in linguistics and psychology, he is active in both the Multi-Agent Systems, the Social Simulation, and the Cognitive Science communities. Program chair of the First International Joint Conference on Autonomous Agents and Multi-Agent Systems – AAMAS-2002; General co-chair of the last International Joint Conference on Autonomous Agents and Multi-Agent Systems – AAMAS-2009; chair of several international workshops in these fields (like ATAL; “Trust and Deception in Artificial Societies”); advisory member of several international conferences and societies (like Cognitive Science; IFMAS); member of the editorial board of "J of Autonomous Agents and MAS", of "Cognitive Science Quarterly", of the MIT CogNet; promoter of the Italian Association for Cognitive Sciences, and of the special interest group on Agents of the AI*IA. Award as “fellows” of the European Coordinating Committee for Artificial Intelligence, for “Pioneering work in the field”; August 2003 PhD Honoris Causa in Cognitive Science at the University of Torino; Award “Mind and Brain” 2008. Univ of Torino. Invited speaker at IJCAI'97 (and many other conferences and workshops in AI, logic, philosophy, linguistics, and psychology, economics).
Research fields of interest include cognitive approach to communication (semantics and pragmatics); cognitive agent theory and architecture; multi-agent systems; agent-based social simulation; social cognition and emotions; cognitive foundations of complex social phenomena (dependence, power, cooperation, norms, organization, social functions, etc.).
Professor Castelfranchi has published more than 200 conference and journal articles on cognitive, computational and formal-theoretical models of social interaction and social mind, and 12 books in Italian.
Cristiano Castelfranchi is Full professor of "Cognitive Sciences" at the University of Siena, Department of Communication Science, and Director of the Institute of Cognitive Sciences and Technologies- ISTC of the National Research Council, in Roma.
A cognitive scientist with a background in linguistics and psychology, he is active in both the Multi-Agent Systems, the Social Simulation, and the Cognitive Science communities. Program chair of the First International Joint Conference on Autonomous Agents and Multi-Agent Systems – AAMAS-2002; General co-chair of the last International Joint Conference on Autonomous Agents and Multi-Agent Systems – AAMAS-2009; chair of several international workshops in these fields (like ATAL; “Trust and Deception in Artificial Societies”); advisory member of several international conferences and societies (like Cognitive Science; IFMAS); member of the editorial board of "J of Autonomous Agents and MAS", of "Cognitive Science Quarterly", of the MIT CogNet; promoter of the Italian Association for Cognitive Sciences, and of the special interest group on Agents of the AI*IA. Award as “fellows” of the European Coordinating Committee for Artificial Intelligence, for “Pioneering work in the field”; August 2003 PhD Honoris Causa in Cognitive Science at the University of Torino; Award “Mind and Brain” 2008. Univ of Torino. Invited speaker at IJCAI'97 (and many other conferences and workshops in AI, logic, philosophy, linguistics, and psychology, economics).
Research fields of interest include cognitive approach to communication (semantics and pragmatics); cognitive agent theory and architecture; multi-agent systems; agent-based social simulation; social cognition and emotions; cognitive foundations of complex social phenomena (dependence, power, cooperation, norms, organization, social functions, etc.).
Professor Castelfranchi has published more than 200 conference and journal articles on cognitive, computational and formal-theoretical models of social interaction and social mind, and 12 books in Italian.
Abstract
I will shortly but systematically examine the crucial problem of goal theory for Agents and MAS. Possibly:
Goal definition and typology (ex. Not all goals are pursued, and a goal is not a goal only when is pursued; not all goals are desires or desirable). The cybernetic and psychological notion of 'goal', and the various functions of a goal (evaluation, action activation and planning, monitoring, adjusting, frustration, satisfaction, etc.)
'Goals' vs. 'Functions': why cognitive coherence is not really a 'goal'; why utility maximization is not really a 'goal';
Terminal vs. instrumental goals; Means-End reversing;
Goal-value; urgency vs. importance;
Goal-dynamics: from 'desires' to 'intentions'; dropping goals;
Motivating vs. non-motivating goals and expectations;
Motives vs. Decisive goals;
Full autonomous agency as Goal-autonomy;
Goal-adoption as the basis for pro-sociality; goal-delegation (reliance and trust)
Embodied goals (needs, true desires, impulses, ...); the felt 'value'; no emotions without goals.
I will shortly but systematically examine the crucial problem of goal theory for Agents and MAS. Possibly:
Goal definition and typology (ex. Not all goals are pursued, and a goal is not a goal only when is pursued; not all goals are desires or desirable). The cybernetic and psychological notion of 'goal', and the various functions of a goal (evaluation, action activation and planning, monitoring, adjusting, frustration, satisfaction, etc.)
'Goals' vs. 'Functions': why cognitive coherence is not really a 'goal'; why utility maximization is not really a 'goal';
Terminal vs. instrumental goals; Means-End reversing;
Goal-value; urgency vs. importance;
Goal-dynamics: from 'desires' to 'intentions'; dropping goals;
Motivating vs. non-motivating goals and expectations;
Motives vs. Decisive goals;
Full autonomous agency as Goal-autonomy;
Goal-adoption as the basis for pro-sociality; goal-delegation (reliance and trust)
Embodied goals (needs, true desires, impulses, ...); the felt 'value'; no emotions without goals.
Getting Agents to Tell the Truth
|
Boi Faltings Ecole Polytechnique Federale de Lausanne (EPFL) Switzerland |
Brief Bio
Boi Faltings is a full professor of computer science at the Ecole Polytechnique Fédérale de Lausanne (EPFL), where he heads the Artificial Intelligence Laboratory. His main research contributions are in the area of qualitative and case-based reasoning, constraint programming, distributed problem-solving, and recommender systems.
He has co-founded 6 companies in e-commerce and computer security and acted as advisor to several other companies world-wide. Prof. Faltings has published over 300 refereed papers and graduated over 25 Ph.D. students, several of which have won national and international awards. Boi Faltings is a fellow of the European Coordinating Committee for Artificial Intelligence. He has served as head of the computer science department from 1996-1998 and as head of the Institute of Core Computing Sciences from 2005-2008. He holds a Diploma from ETH Zurich and a Ph.D. from the University of Illinois at Urbana-Champaign.
Boi Faltings is a full professor of computer science at the Ecole Polytechnique Fédérale de Lausanne (EPFL), where he heads the Artificial Intelligence Laboratory. His main research contributions are in the area of qualitative and case-based reasoning, constraint programming, distributed problem-solving, and recommender systems.
He has co-founded 6 companies in e-commerce and computer security and acted as advisor to several other companies world-wide. Prof. Faltings has published over 300 refereed papers and graduated over 25 Ph.D. students, several of which have won national and international awards. Boi Faltings is a fellow of the European Coordinating Committee for Artificial Intelligence. He has served as head of the computer science department from 1996-1998 and as head of the Institute of Core Computing Sciences from 2005-2008. He holds a Diploma from ETH Zurich and a Ph.D. from the University of Illinois at Urbana-Champaign.
Abstract
There are many scenarios where we would like agents to report their observations or expertise in a truthful way.
Game-theoretic principles can be used to provide incentives to do so.
I survey several approaches to eliciting truthful information, in particular scoring rules, peer prediction methods and opinion polls.
There are many scenarios where we would like agents to report their observations or expertise in a truthful way.
Game-theoretic principles can be used to provide incentives to do so.
I survey several approaches to eliciting truthful information, in particular scoring rules, peer prediction methods and opinion polls.
Uncertainty Theories, Degrees of Truth and Epistemic States
|
Didier Dubois Institut de Recherche en Informatique de Toulouse (IRIT) France |
Brief Bio
Didier Dubois is a Research Advisor at IRIT, the Computer Science Department of Paul Sabatier University in Toulouse, France and belongs to the French National Centre for Scientific Resarch (CNRS). His topics of interest range from Artificial Intelligence to Operations Research and Decision Sciences, with emphasis on the modelling, representation and processing of imprecise and uncertain information in reasoning, decision and risk analysis.
He is the co-author, with Henri Prade, of two monographs on fuzzy sets and possibility theory, and 15 edited volumes on uncertain reasoning, fuzzy sets, and decision analysis. Also with Henri Prade, he coordinated the HANDBOOK of FUZZY SETS series published by Kluwer (7 volumes, 1998-2000) including the book Fundamentals of Fuzzy Sets. He has contributed about 200 technical journal papers on uncertainty theories and applications. He is a Co-Editor-in-Chief of the journal Fuzzy Sets and Systems and a member of the Editorial Board of several technical journals dealing with uncertain reasoning. He is a former president of the International Fuzzy Systems Association (IFSA, 1995-1997) and an IFSA and an ECCAI fellow. He received the 2002 Pioneer Award of the IEEE Neural Network Society.
Didier Dubois is a Research Advisor at IRIT, the Computer Science Department of Paul Sabatier University in Toulouse, France and belongs to the French National Centre for Scientific Resarch (CNRS). His topics of interest range from Artificial Intelligence to Operations Research and Decision Sciences, with emphasis on the modelling, representation and processing of imprecise and uncertain information in reasoning, decision and risk analysis.
He is the co-author, with Henri Prade, of two monographs on fuzzy sets and possibility theory, and 15 edited volumes on uncertain reasoning, fuzzy sets, and decision analysis. Also with Henri Prade, he coordinated the HANDBOOK of FUZZY SETS series published by Kluwer (7 volumes, 1998-2000) including the book Fundamentals of Fuzzy Sets. He has contributed about 200 technical journal papers on uncertainty theories and applications. He is a Co-Editor-in-Chief of the journal Fuzzy Sets and Systems and a member of the Editorial Board of several technical journals dealing with uncertain reasoning. He is a former president of the International Fuzzy Systems Association (IFSA, 1995-1997) and an IFSA and an ECCAI fellow. He received the 2002 Pioneer Award of the IEEE Neural Network Society.
Abstract
Uncertainty has been addressed by two often disjoint communities: probability-based and logic-based artificial intelligence.
On the side of Bayesianism, there is a questionable dogma pertaining to the universal capability of single probability distributions to represent any state of information. In fact, probability theory more naturally captures ideas of repeatable experiments. On the side of logics, the epistemic state of an agent is often merely represented as a subset of possible worlds, thus capturing the idea of incomplete knowledge. New uncertainty theories are putting the two dimensions together by considering random epistemic states or subsets of probabilities.
However, uncertainty has not really received much attention in the tradition of logic. The interaction between uncertainty and truth even seems to have created some confusion. Historically, there is a temptation to extend the truth-set underlying a given logic with values expressing epistemic notions such as ignorance and contradiction. It then leads to some truth-functional many-valued logic different from classical logic, but often using the same syntax. This is the case for instance with original Kleene or Lukasiewicz logics, partial logic and Belnap bilattice logic.
This talk insists that neither ignorance nor contradiction can be viewed as additional truth-values nor processed in a truth-functional manner, and that doing it leads to weak or debatable uncertainty handling approaches. Modal logics have been used to handle epistemic notions in a more satisfactory way, but the Kripke semantics based on relations over possible worlds are perhaps unnecessarily complex to handle plain notions of epistemic states. A simpler semantics for representing an agent's knowledge about another agent's beliefs can be obtained by moving from possible worlds to possible epistemic states.
We suggest that, in order to handle such epistemic notions at the language level, we need two-tiered systems where one logic (typically propositional logic) describing events is encapsulated by another one (that can be many-valued) describing beliefs as testimonies received by an agent. In the multi-agent case, this approach leads to a more satisfactory handling of inconsistency close to Kyburg's logic of risky knowledge and monotonic modal logics where the adjunction law does not hold.
Such an approach paves the way to a general approach to more expressive logics of uncertainty, along the lines suggested by Esteva, Godo and Hajek casting probability, possibility or belief functions within a suitable multiple-valued setting, where beliefs are graded, but events remain Boolean.
Uncertainty has been addressed by two often disjoint communities: probability-based and logic-based artificial intelligence.
On the side of Bayesianism, there is a questionable dogma pertaining to the universal capability of single probability distributions to represent any state of information. In fact, probability theory more naturally captures ideas of repeatable experiments. On the side of logics, the epistemic state of an agent is often merely represented as a subset of possible worlds, thus capturing the idea of incomplete knowledge. New uncertainty theories are putting the two dimensions together by considering random epistemic states or subsets of probabilities.
However, uncertainty has not really received much attention in the tradition of logic. The interaction between uncertainty and truth even seems to have created some confusion. Historically, there is a temptation to extend the truth-set underlying a given logic with values expressing epistemic notions such as ignorance and contradiction. It then leads to some truth-functional many-valued logic different from classical logic, but often using the same syntax. This is the case for instance with original Kleene or Lukasiewicz logics, partial logic and Belnap bilattice logic.
This talk insists that neither ignorance nor contradiction can be viewed as additional truth-values nor processed in a truth-functional manner, and that doing it leads to weak or debatable uncertainty handling approaches. Modal logics have been used to handle epistemic notions in a more satisfactory way, but the Kripke semantics based on relations over possible worlds are perhaps unnecessarily complex to handle plain notions of epistemic states. A simpler semantics for representing an agent's knowledge about another agent's beliefs can be obtained by moving from possible worlds to possible epistemic states.
We suggest that, in order to handle such epistemic notions at the language level, we need two-tiered systems where one logic (typically propositional logic) describing events is encapsulated by another one (that can be many-valued) describing beliefs as testimonies received by an agent. In the multi-agent case, this approach leads to a more satisfactory handling of inconsistency close to Kyburg's logic of risky knowledge and monotonic modal logics where the adjunction law does not hold.
Such an approach paves the way to a general approach to more expressive logics of uncertainty, along the lines suggested by Esteva, Godo and Hajek casting probability, possibility or belief functions within a suitable multiple-valued setting, where beliefs are graded, but events remain Boolean.
The MIT Deliberatorium: Enabling Large-Scale Deliberation about Complex Systemic Problems
|
Mark Klein MIT Center for Collective Intelligence United States of America |
Brief Bio
Dr. Mark Klein (http://cci.mit.edu/klein/) is a Principal Research Scientist at the MIT Center for Collective Intelligence, as well as an Affiliate at the MIT Computer Science and AI Lab (CSAIL) and the New England Complex Systems Institute (NECSI). His research focuses on understanding how computer technology can help groups, especially large ones, make better decisions about complex problems. He has made contributions in the areas of computer-supported conflict management for collaborative design, design rationale capture, business process re-design, exception handling in workflow and multi-agent systems, service discovery, negotiation algorithms, 'emergent' dysfunctions in distributed systems and, more recently, 'collective intelligence' systems to help people collaboratively solve complex problems like global warming.
Dr. Mark Klein (http://cci.mit.edu/klein/) is a Principal Research Scientist at the MIT Center for Collective Intelligence, as well as an Affiliate at the MIT Computer Science and AI Lab (CSAIL) and the New England Complex Systems Institute (NECSI). His research focuses on understanding how computer technology can help groups, especially large ones, make better decisions about complex problems. He has made contributions in the areas of computer-supported conflict management for collaborative design, design rationale capture, business process re-design, exception handling in workflow and multi-agent systems, service discovery, negotiation algorithms, 'emergent' dysfunctions in distributed systems and, more recently, 'collective intelligence' systems to help people collaboratively solve complex problems like global warming.
Abstract
Humanity now finds itself faced with highly complex challenges – ranging from climate change, and the spread of disease to international security and scientific collaborations - that require effective collective multi-disciplinary decision making with large communities that are distributed in time and space. While social computing tools (e.g. web forums, wikis, email, instant messaging, media sharing sites, social networking, and so on) have created unprecedented opportunities for connecting and sharing on a massive scale, they still fare poorly when applied to deliberation, i.e. the systematic identification, evaluation, and convergence upon solutions to complex problems. Internet-mediated discussions are instead all-to-often characterized by widely varying quality, poor signal-to-noise ratios, spotty coverage, and scattered content, as well as dysfunctional dynamics for controversial issues.
This talk will present a novel integration of ideas taken from social computing and argumentation theory that has, we believe, the potential to address these limitations. We will describe the underlying concepts, the results of our evaluations to date, and some promising directions for future work.
Humanity now finds itself faced with highly complex challenges – ranging from climate change, and the spread of disease to international security and scientific collaborations - that require effective collective multi-disciplinary decision making with large communities that are distributed in time and space. While social computing tools (e.g. web forums, wikis, email, instant messaging, media sharing sites, social networking, and so on) have created unprecedented opportunities for connecting and sharing on a massive scale, they still fare poorly when applied to deliberation, i.e. the systematic identification, evaluation, and convergence upon solutions to complex problems. Internet-mediated discussions are instead all-to-often characterized by widely varying quality, poor signal-to-noise ratios, spotty coverage, and scattered content, as well as dysfunctional dynamics for controversial issues.
This talk will present a novel integration of ideas taken from social computing and argumentation theory that has, we believe, the potential to address these limitations. We will describe the underlying concepts, the results of our evaluations to date, and some promising directions for future work.
Model Driven Design of Agents and Multiagent Systems
|
Klaus Fischer Agents and Simulated Reality, DFKI GmbH Germany |
Brief Bio
Klaus Fischer is research fellow and head of the multiagent system group in the department of Agents and Simulated Realities at the German Research Center for Artificial Intelligence (DFKI) in Saarbrücken. He holds a doctoral degree and a diploma for Computer Science from Technische Universität (TU) in München.
His main research interest is in model driven design of multiagent systems with applications in virtual enterprises, supply chain and logistics management, and virtual and simulated realities. Most recent achievements of the MAS group have been the development of the domain specific modeling language DSML4MAS with the metamodel PIM4Agents in its core and the development of an agent-based shop floor control system for steel production which is running 24/7 in the steel work of Saarstahl in Völklingen. Further important contributions of the group were in the area of the Semantic Web and semantic reasoning.
Klaus Fischer has led a significant number of research projects (e.g. EC projects ATHENA, COIN, and SHAPE) and industrial project (most recently with Saarstahl AG). He actively contributes to organizing international scientific events (e.g. Co-program chair IAT 2011, SPC member of AAMAS 2010) and further serves in the editorial boards of the Journal of Autonomous Agents and Multiagent Systems (JAAMAS) and Applied Intelligence (APIN).
Klaus Fischer is research fellow and head of the multiagent system group in the department of Agents and Simulated Realities at the German Research Center for Artificial Intelligence (DFKI) in Saarbrücken. He holds a doctoral degree and a diploma for Computer Science from Technische Universität (TU) in München.
His main research interest is in model driven design of multiagent systems with applications in virtual enterprises, supply chain and logistics management, and virtual and simulated realities. Most recent achievements of the MAS group have been the development of the domain specific modeling language DSML4MAS with the metamodel PIM4Agents in its core and the development of an agent-based shop floor control system for steel production which is running 24/7 in the steel work of Saarstahl in Völklingen. Further important contributions of the group were in the area of the Semantic Web and semantic reasoning.
Klaus Fischer has led a significant number of research projects (e.g. EC projects ATHENA, COIN, and SHAPE) and industrial project (most recently with Saarstahl AG). He actively contributes to organizing international scientific events (e.g. Co-program chair IAT 2011, SPC member of AAMAS 2010) and further serves in the editorial boards of the Journal of Autonomous Agents and Multiagent Systems (JAAMAS) and Applied Intelligence (APIN).
Abstract
The lecture presents an innovative approach to the design of multiagent systems. It explains how the basic ideas behind the model driven architecture (MDA) proposed by the Object Management Group (OMG) can be successfully translated into a framework for the design of agents and multiagent systems and presents the domain-specific modeling language DSML4MAS as a concrete example of such a framework. Although DSML4MAS was originally applied to the design of multiagent systems in the environment of service-oriented architectures in the context of business applications in virtual enterprises, DSML4MAS is more recently also adopted for the design of agents in simulated realities. The lecture focuses on the general idea behind MDA and presents details of DSML4MAS as well as examples in the different application areas.
The lecture presents an innovative approach to the design of multiagent systems. It explains how the basic ideas behind the model driven architecture (MDA) proposed by the Object Management Group (OMG) can be successfully translated into a framework for the design of agents and multiagent systems and presents the domain-specific modeling language DSML4MAS as a concrete example of such a framework. Although DSML4MAS was originally applied to the design of multiagent systems in the environment of service-oriented architectures in the context of business applications in virtual enterprises, DSML4MAS is more recently also adopted for the design of agents in simulated realities. The lecture focuses on the general idea behind MDA and presents details of DSML4MAS as well as examples in the different application areas.