


Context in Knowledge Representation (KR) and Natural Language (NL)

Context plays a crucial role in human knowledge representation, reasoning, natural language processing, and perception. Thus,
computer systems which act "intelligently" need the ability to represent, utilize and reason about contexts.

Within the AI community, high hopes were set for ways in which context could help reasoning systems. A number of projects
aimed at incorporating the notion of context into intelligent KR and NL computer systems were initiated. However, this task
turned out to be more difficult than originally anticipated. The accepted view now seems to be that a better understanding of
the phenomenon of context and its general mathematical and computational properties are needed before such systems can
fully cash in on the "magic of context."

Making progress in this direction will require an interdisciplinary collaboration. Other academic disciplines, such as linguistics,
philosophy and anthropology, have also studied various aspects of the context phenomena. However, their theories usually lie
embedded in the analysis of specific linguistic constructions, so applying them in AI systems is a research challenge in and of
itself.

This symposium aims at bringing together researchers in AI, linguistics, philosophy, cognitive science, and other related fields
who are interested in theoretical and practical aspects of context.

Some technical issues of interest are:

     What are some roles of context in KR and NL systems, particularly in the process of reasoning? We are primarily
     interested in formal, computable theories addressing the roles of context. 
     Is it possible to decrease the computational complexity of a formal system by the means of introducing context? 
     What is context? 
     Is context an inherent characteristic of natural language that ultimately decides the formal power of natural language? 
     How can we represent relations between contexts? How can a computer system automatically infer the relation between
     some given set of contexts? 
     Is decontextualization possible/necessary? 
     How can information obtained in one context be utilized in another, possibly unanticipated, context? 
     Which aspects of context or which contexts result in refined, more general, and different interpretations of natural
     language? 
     Do the existing theories of context in KR offer insights or solutions into the context dependency of natural language? 
     Do natural language theories which involve context offer solutions to the problems which have motivated the
     development of theories of context in KR? 
     Which aspects of context can be handled in a real-life application such as managing a large knowledge base or
     processing large volumes of text?

This list is not exhaustive and papers on any topic concerning context in KR or NL are welcome.

Participants: This is a double-size, about 80-100 participant symposium.

Submission Information

Hard copy submissions should be no longer than 12 pages in a 12 point font. Please send 6 hard copies of your paper to:

     Context Fall Symposium
     American Association for Artificial Intelligence
     445 Burgess Drive Menlo Park, CA 94025-3442

In addition please place the PostScript version of your paper in the "~pub/iwanska/context" directory via the anonymous ftp at:
ftp.cs.wayne.edu.

Organizing Committee

Sasa Buvac (cochair), Stanford University (buvac@cs.stanford.edu); Lucja Iwanska (cochair), Wayne State University
(lucja@cs.wayne.edu); Kees van Deemter, Philips Electronics (deemter@natlab.research.philips.com); Fausto Giunchiglia,
IRST & Universit di Trento (fausto@irst.itc.it); R.V. Guha, Apple Computers, Inc. (guha@taurus.apple.com); Pat Hayes,
University of Illinois (phayes@picayune.coginst.uwf.edu); Graeme Hirst, University of Toronto (gh@cs.toronto.edu); John
McCarthy, Stanford University (jmc@cs.stanford.edu); Stuart Shapiro, SUNY Buffalo (shapiro@cs.buffalo.edu); Rich
Thomason, University of Pittsburgh (thomason@isp.pitt.edu); Wlodek Zadrozny, IBM TJ Watson Research Center
(wlodz@watson.ibm.com).
