RBFCM Based Heuristic Coordinator Algorithm for KDD

Abstract

Heuristic coordinator, as an import part in KDD systems based on double bases cooperating mechanism, can simulate “creating intent” of cognitive psychology feature and enhance the ability of self-cognition. With the aim to improve the cognitive feature and the performance of heuristic coordinator, the paper proposes one new heuristic coordinator algorithm, which uses rule based fuzzy cognitive map to represent knowledge and to be effective inference in order to get non-association state of knowledge base for directional mining shortage knowledge in massive database. And the experiment demonstrates that the approach effectively reduces the searching space and increases the intelligence of knowledge discovery compared with the directed hyper-graph based method.