Developing a GP-based Framework for Knowledge Integration

Abstract

Knowledge integration is one of the important tasks for applying knowledge management in an organization to improve organizational performance and competitive competence. In this paper, we have proposed a GP-based knowledge-integration framework that automatically combines multiple rule sets into one integrated knowledge base. The proposed framework consists of three phases: knowledge collection and translation, knowledge integration, and knowledge output. In the collection and translation phase, each knowledge source is obtained and expressed as a rule set and then translated as a classification tree. In the integration phase, the genetic programming technique is used to generate a nearly optimal classification tree. In the output phase, the final derived classification tree is transferred as a rule set, then output to the knowledge base to facilitate the inference of new data. Two new genetic operators, abridgement and compromise, are designed in the proposed approach to remove redundancy, subsumption and contradiction, thus producing more accurate and concise classification rules than that without using them. Experimental results from diagnosis of breast cancer also show the feasibility of the proposed algorithm.