This paper presents the issues of dealing with context from the perspective of cognitive modeling. A dynamic theory of context is proposed which considers context as the set of all entities that influence human cognitive behavior on a particular occasion. As a consequence context is thought of as the dynamic fuzzy set of all associatively relevant memory elements at a particular instant of time. These memory elements might be both mental representations and operations. Some experimental facts about the influence of the perceptible environment as well as of the previous memory state on human problem solving are briefly presented. The dynamic nature of context influence on behavior is emphasized. A general cognitive architecture, DUAL, is presented which consists of many small agents running autonomously in parallel with variable speeds depending on their current associative relevance. A model of problem solving, AMBR, based on DUAL is discussed where problem solving emerges from the collective behavior of the agents. The possibilities of AMBR for modeling context and priming effects are considered and some simulation results are presented.
[1]
Gary G. Hendrix,et al.
Encoding Knowledge in Partitioned Networks
,
1979
.
[2]
D. Sperber,et al.
Relevance: Communication and Cognition
,
1997
.
[3]
W. Kintsch.
The role of knowledge in discourse comprehension: a construction-integration model.
,
1988,
Psychological review.
[4]
D. Sperber,et al.
Relevance: Communication and Cognition
,
1989
.
[5]
Boicho Kokinov,et al.
Associative Memory-Based Reasoning: Some Experimental Results
,
1990
.
[6]
Boicho N. Kokinov,et al.
The DUAL Cognitive Architecture: A Hybrid Multi-Agent Approach
,
1994,
ECAI.
[7]
Melanie Mitchell,et al.
The Copycat project: a model of mental fluidity and analogy-making
,
1995
.
[8]
Douglas R. Hofstadter,et al.
Fluid Concepts and Creative Analogies
,
1995
.
[9]
Laurent Perrussel,et al.
Contextual Reasoning
,
1998,
ECAI.
[10]
Irene Santini de Sigel,et al.
Communication and Cognition
,
1998
.
[11]
James A. Anderson.
Associative networks
,
1998
.
[12]
Boicho Kokinov,et al.
Context Effects on Problem Solving
,
1999
.