A self-organizing pattern retrieval system and its applications

Abstract

This article is an overview of research on the design and application of a Self-organizing Pattern Retrieval System. This system is a nontraditional form of database for several reasons: (1) The objects (patterns) stored in the database are complex data structures such as graphs, or matrices as opposed to relations, sets, lists or frames; (2) the system attempts to optimize its internal structure for retrieval efficiency by taking advantage of classificatory information (common patterns) it discovers about the patterns in the database; and (3) the system is representation-independent, allowing the user to use whatever representation scheme he desires as long as he provides a small set of I/O routines and comparison functions for the representation. This system has been applied to the organization of molecules and reactions in organic chemistry. The system is able to select valid precursors to molecules based on its knowledge of real-world reactions. It is also able to suggest new chemical reactions based on generalizations from its knowledge base. The system has also been applied to the retrieval, generalization, and classification of chess positions. Its ability to recognize the tactical similarity of positions is compared with that of a chess master. The design objectives of a system for classifying Radio Frequency Interference required by NASA's SET1 (Search for ExtraTerrestrial Intelligence) project is also discussed.

A lattice conceptual clustering system and its application to browsing retrieval

“…Another feature of concept lattices is that they can generate overlapping classes. Although this is not a new idea (e.g., Levinson, 1984;Lebowitz, 1986) the importance of this issue has been somewhat overlooked. On one hand, very little work has been done on acquiring formally characterizable classes of partially ordered sets of clusters;…”

Approaches to Parallel Graph-Based Knowledge Discovery

“…Background knowledge can take the form of known substructure models that may potentially appear in the database or graph match rules to adjust the cost of each inexact graph match test. Unlike other existing approaches to graph-based discovery [5,19,25,28,28], Subdue is effective at finding interesting and repetitive substructures in any structural database with or without domain-specific guidance. The results of the scalability study in Section 3 are demonstrated on databases in two different domains.…”

Untitled

“…Although this is not a new idea (e.g. (Levinson, 1984;Lebowitz, 1986)) the importance of this issue has been somewhat overlooked. On one hand, very little work has been done on acquiring formally characterizable classes of partially ordered sets of clusters; on the other hand, although learning overlapping classes may be an important practical requirement even for pure predictive tasks (Martin & Billman, 1994), data sets with non-disjoint classes have been surprisingly rare in the machine learning literature.…”