Abstract Syntactical pattern recognition techniques are applied to conceptual patterns for analogical problem solving. The conceptual patterns are concerned with relationships between entities. A problem to be solved (or a question to be answered) through analogical reasoning is treated as an incomplete pattern (in the target domain) with some missing information. Given a pattern base (which consists of known conceptual patterns) and a target pattern with missing information, the task of analogical problem solving is to recognize a known pattern (in some source domain) in the pattern base which contains a portion similar to the known portion of the target pattern, so that the missing information in the target domain can be fulfilled through the known pattern through structure mapping. In the context of conceptual pattern recognition, the structure mapping process is performed through mapping of grammar. Some important terms are defined, a grammar for pattern description is described, and the function of a grammar mapping engine (GME) is then discussed. A brief comparison with related work is also provided.
[1]
Robert J. Schalkoff,et al.
Pattern recognition : statistical, structural and neural approaches / Robert J. Schalkoff
,
1992
.
[2]
Sing-Tze Bow,et al.
Pattern recognition and image preprocessing
,
1992
.
[3]
Siegfried Treu.
Recognition of Logical Interface Structures to Enhance Human-Computer Interaction
,
1989,
Int. J. Pattern Recognit. Artif. Intell..
[4]
Peter P. Chen.
The Entity-Relationship Model: Towards a unified view of Data
,
1976
.
[5]
Dedre Gentner,et al.
Transfer of training as analogical mapping
,
1988,
IEEE Trans. Syst. Man Cybern..
[6]
Dedre Gentner,et al.
Analogical Inference and Analogical Access.
,
1987
.
[7]
Dan W. Patterson,et al.
Introduction to artificial intelligence and expert systems
,
1990
.
[8]
Julius T. Tou,et al.
Pattern Recognition Principles
,
1974
.
[9]
Brian Falkenhainer,et al.
The Structure-Mapping Engine: Algorithm and Examples
,
1989,
Artif. Intell..