Toolkit for Agent-based Knowledge Extraction (TAKE)

摘要

This paper describes the motivations and concept of operation of a Toolkit for Agent-bused Knowledge Extraction (TAKE). TAKE implements a hybrid approach that combines emerging knowledge discovery techniques with traditional ontology engineering methods to extract knowledge from emerging distributed and disparate data sources and rapidly deploy this knowledge for the analysis, reengineering, and redesign of complex multi-agent systems. TAKE provides a comprehensive framework for knowledge management that encompasses the overall knowledge management process starting with the definition of a problem to the delivery of (ⅰ) structured knowledge to an ontology/knowledge base and (ⅱ) valuable systems reengineering / redesign assessment information. TAKE research contributions include (ⅰ) a hybrid approach to combine soft computing techniques with ontology engineering methods for automated knowledge extraction and analysis, (ⅱ) innovative self-adaptation mechanisms that ensure automated revisions to knowledge haxes and knowledge analysis strategies in response to rapidly changing application environments, and (ⅲ) a scalable and re-configurable distributed agent architecture that enables rapid and cost effective knowledge based systems reengineering and redesign.