Seventh International Workshop on Nonmonotonic Reasoning


Trento, Italy
May 30 - June 1, 1998

in conjunction with the

Sixth International Conference on Principles of Knowledge Representation and Reasoning (KR'98)

with support from AAAI, Compulog, and Associazione Italiana per l'Intelligenza Artificiale (AI*IA)

The aim of the workshop is to bring together active researchers interested in the area of nonmonotonic reasoning to discuss current research, results, and problems of both a theoretical and practical nature. This time the workshop consists of plenary sessions and specialized workshops on the following topics:

Program

List of invited speakers: See the program for more details.

Workshop Location

NM'98 will be held in the Facolta` di Lettere e Filosofia, third floor, rooms 7/S, 8/S, 9/S. These are located at the Centro Servizi Culturali S. Chiara, Via S. Croce 67, 38100 Trento (marked KR'98 on the map). Signs will direct people once in S. Chiara.

The NM'98 conference reception desk will be located in front of the rooms from 8.45am to 9.15am and during the coffee breaks of the first day.

Submission and Participation

Submission of papers is through the specialized workshops. Please consult the details for the particular workshops via the links above. The deadline for submissions is Feb 23, 1998.

Participation in the workshops will be by invitation only. Those wishing to attend without submitting a paper (e.g. students) should send a statement of interest with a short description of research interests and recent work to one of the co-chairs.

Proceedings

Each specialized workshop distributes the accepted papers at the meeting (details available for each of the workshops through the respective links above). It is planned to collect a proceedings volume of the whole workshop from selected papers presented at the specialized workshops.

Workshop Chairs

Ray Reiter (Honorary Chair)
Gerd Brewka (Co-chair)
Enrico Giunchiglia (Local Chair)
Ilkka Niemelä (Co-chair)

Student support

AAAI is providing travel support for a limited number of participating students. Preference will be given to students presenting a paper and to students from overseas. Applications for student support should be sent electronically to one of the co-chairs. They should include address of home university, estimation of travel costs and, if applicable, title of paper presented at NM'98.

Further Information

Gerd Brewka (Co-chair)
Universitaet Leipzig
Institut fuer Informatik
Augustusplatz 10/11
04109 Leipzig, Germany
brewka@informatik.uni-leipzig.de
phone: +49 341 97 32235
fax: +49 341 97 32299
Ilkka Niemelä (Co-chair)
Helsinki University of Technology
Dept. of Computer Science and Engineering
Digital Systems Laboratory
P.O.Box 1100, FIN-02015 HUT, Finland
Ilkka.Niemela@hut.fi
phone: +358-9-451 3290
fax: +358-9-465 077

Abstracts of the Invited Talks

From Logic Programming to Multi-agent Systems
Bob Kowalski, Imperial College

Abstract: Some years ago, some of us thought that logic programming might be sufficiently powerful to serve as a new foundation for all of Computing, unifying such otherwise diverse areas as programming, specification, databases, and knowledge representation. It is now obvious that this early promise has not borne fruit, and it is timely to ask whether there any essential technical reasons for this failure.

Recent work on extending logic programming to multi-agent systems can be viewed as revealing a number of fundamental deficiencies in the simple model of logic programing. In the simple model, logic programs can be regarded as the "rational", goal-reduction component of a single agent. To be more comprehensive, even in the single agent case, logic programs need to be extended, to be reactive to changes that arise in the environment. They also need to be extended to the multi-agent case. Both extensions can be made gracefully and are needed for other purposes.

In this talk, I will describe an architecture for multi-agent systems that achieves reactivity, by employing integrity constraints in addition to logic programs. It achieves the ability for agents to interact concurrently with other agents and with their environment, by treating all interactions as interactions between an agent and a global environment. In this respect, concurrency is achieved in the manner of blackboard architectures and Linda-like systems rather than in the manner of message passing architectures, as incorporated, for example in concurrent logic programming.

Pragmatism, Inquiry and Inductively Extended Expansion
Isaac Levi, Columbia University

Abstract: The Belief - Doubt Model of inquiry proposed by Charles Peirce and supported by William James and John Dewey insists that justification is not needed for current beliefs but only for changing beliefs (whether by expansion or contraction). Hence, the classical pragmatists are the pioneers in studying justifying belief change rather than justifying current beliefs as Cartesian foundationalists and coherentists alike are concerned to do. Pragmatists also believe that justification of change of belief requires showing that such changes best promote the goals of inquiry. For Peirce, at least, the goals of specific inquiries are cognitive goals not reducible to practical, moral, economic, political, aesthetic aims. I suggest that such goals always ought to be concerned with obtaining new valuable information and avoiding error.

In this discussion, I illustrate this point of view by rehearsing the account of inductive expansion I have been proposing since the 1960's and explore the relations between inductive expansion and the view of ampliative, nonmonotonic reasoning based on the theory of expectations discussed by Gärdenfors and Makinson.

Nonmonotonic Reasoning in the Service of Elaboration Tolerance
John McCarthy, Stanford University

Abstract: My first paper on circumscription mentioned eliminating the possibility of there being a bridge in the missionaries and cannibals problem by circumscription. The problem doesn't mention a bridge, so the puzzle solver is to infer than none exists. The nonmonotonic formalisms have been tested by verifying that they indeed eliminate unwanted possibilities. However, there is a converse problem. An elaboration tolerant formalism would admit adding a statement about there being a bridge. The formalisms also need to be tested that they have reasonable consequences when new facts are added. Thus if a bridge is mentioned, the solver should be able to verify a plan that uses the bridge while still being able to infer the non-existence of a helicopter. The lecture will discuss what conditions elaboration tolerance puts on nonmonotonic reasoning formalisms and on axiomatizations of particular domains.

Learning to Make Nonmonotonic Inferences
Dan Roth, University of Illinois at Urbana/Champaign

Abstract: While the central role of learning in cognition is acknowledged by many, most lines of research nevertheless study reasoning phenomena separately from learning phenomena. In addition to computational difficulties, this leads to methodological problems that arise from the need to take a uniform view of how to treat incomplete information.

We will describe an approach that views reasoning and, in particular, handling incomplete information and the nonmonotonic phenomena as an inductive phenomena. One of the main premises that underlie this work is that missing information in the interaction with the environment may be as informative for future interactions as observed information.

The computational and methodological advantages of the ``Learning to Reason'' approach will be discussed and illustrated using a few ``classical'' nonmonotonic examples and some large-scale examples of real-world inferences in the natural language domain.

From Features and Fluents to Thinking when Flying -
Reasoning about Actions in an Intelligent UAV
Erik Sandewall, Linköping University

Abstract: Unmanned Aerial Vehicles (UAV's) offer an interesting challenge for intelligent robotic systems, and in particular for reasoning about actions and change. The UAV itself needs to deliberate on its maneuvering, knowledge acquisition, and communication actions. Depending on the type of application of the UAV, it may also have to deliberate on the actions of fellow UAV's, vehicles or other devices that it observes on the ground, and communication partners.

In this talk I will discuss how our approach to reasoning about actions and change has been (1) utilized and (2) influenced by our participation in the WITAS project, which is an ongoing, large research effort for the development of information technologies for autonomous and intelligent UAV's.