Researches On Dynamic Participation Knowledge Management System Based On Task-Circumstance

The basic goal of Knowledge Management (KM) is promoting knowledge sharing. The success for this will be embodied mainly in using valuable knowledge into work practice. Since knowledge is often implicit in the content of information bodies, how to smartly focus the attention of knowledge workers and search engines on the information bodies valuable for current work practice has becomes the hot point of KM research and the challenging problem urgent to solve.Under such background, this paper proposes a method, called TCKM (Task-Circumstance-based KM), which adopts the structural description oriented to task circumstance as the semantic indexes for the suitability of information body content and the query requirement of work practice, hence can solve effectively this problem. Also, TCKM will support the development and execution of KM systems with high performance from two dimensions: constructing shared understanding foundation, promoting the open evolution of OM (Organization Memory), and in two levels: knowledge process and knowledge meta-process.Using Ontology theory in the field of KM can be called KM Ontology. In this paper we mainly researched on these factions of KM Ontology: content, methods of formalization, maintenance and creation.The chief condition of realizing the dynamic participation of KM is constructing shared understanding foundation, which mainly includes two tasks: one is the creation of basic term system, which can be understood by all KM workers; the other is providing tools and supported environment, which can assist KM workers to construct their term system.The open integration of Knowledge depends on knowledge conversion, knowledge extraction and knowledge stanchion. The evolution of OM and WEB Knowledge extraction effectively promote open integration of knowledge. This paper introduces a method of knowledge extraction, which is based on Ontology. The method uses the hiberarchy of extracted information as information extraction root, and uses induction study to produce rules of information extraction, and is effectively used in extracting WEB information.