TY - JOUR
T1 - Defaults and relevance in model-based reasoning
AU - Khardon, Roni
AU - Roth, Dan
N1 - Funding Information:
Keywords: Knowledge representation; Common-sense reasoning; Learning to reason; Reasoning with models; Context; Default reasoning * Corresponding author. Email: roni@das.harvard.edu. Research supported by AR0 under grant DAALOS-92-G-01 15 and by ONR grant NOOOl4-96-I-0550. ’ An earlier version of the paper appears in Proceedings of the International Joint Conference on Art$icial Intelligence (IJCAI-95) ?E mail: danr@wisdom.weizmann.ac.il. Research supported by the Feldman Foundation. Part of this work was done while at Harvard University, supported by NSF grant CCR-92-00884, DARPA AFOSR-F4962-92-J-0466 and ONR grant NOOO14-96-1-0550.
PY - 1997/12
Y1 - 1997/12
N2 - Reasoning with model-based representations is an intuitive paradigm, which has been shown to be theoretically sound and to possess some computational advantages over reasoning with formula-based representations of knowledge. This paper studies these representations and further substantiates the claim regarding their advantages. In particular, model-based representations are shown to efficiently support reasoning in the presence of varying context information, handle efficiently fragments of Reiter's default logic and provide a useful way to integrate learning with reasoning. Furthermore, these results are closely related to the notion of relevance. The use of relevance information is best exemplified by the filtering process involved in the algorithm developed for reasoning within context. The relation of defaults to relevance is viewed through the notion of context, where the agent has to find plausible context information by using default rules. This view yields efficient algorithms for default reasoning. Finally, it is argued that these results support an incremental view of reasoning in a natural way, and the notion of relevance to the environment, captured by the Learning to Reason framework, is discussed.
AB - Reasoning with model-based representations is an intuitive paradigm, which has been shown to be theoretically sound and to possess some computational advantages over reasoning with formula-based representations of knowledge. This paper studies these representations and further substantiates the claim regarding their advantages. In particular, model-based representations are shown to efficiently support reasoning in the presence of varying context information, handle efficiently fragments of Reiter's default logic and provide a useful way to integrate learning with reasoning. Furthermore, these results are closely related to the notion of relevance. The use of relevance information is best exemplified by the filtering process involved in the algorithm developed for reasoning within context. The relation of defaults to relevance is viewed through the notion of context, where the agent has to find plausible context information by using default rules. This view yields efficient algorithms for default reasoning. Finally, it is argued that these results support an incremental view of reasoning in a natural way, and the notion of relevance to the environment, captured by the Learning to Reason framework, is discussed.
KW - Common-sense reasoning
KW - Context
KW - Default reasoning
KW - Knowledge representation
KW - Learning to reason
KW - Reasoning with models
UR - http://www.scopus.com/inward/record.url?scp=0031359622&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0031359622&partnerID=8YFLogxK
U2 - 10.1016/s0004-3702(97)00044-1
DO - 10.1016/s0004-3702(97)00044-1
M3 - Article
AN - SCOPUS:0031359622
SN - 0004-3702
VL - 97
SP - 169
EP - 193
JO - Artificial Intelligence
JF - Artificial Intelligence
IS - 1-2
ER -