Research Of An Information Retrieval Model For Ontology-Based Enterprise Knowledge Management System

Knowledge management and knowledge management system have recently been paid many attentions. Particularly enterprise knowledge management system(EKMS) oriented to the enterprise application is playing more and more important role. For enterprise knowledge base is becoming huge and complicated, it is a quite challenging task to correctly find the required knowledge in the massive knowledge base. But it is very meaningful especially for the enterprise's efficiency. Hence, how to help employees to search information more efficiently is the key point of current research in information retrieval engines for EKMS. However, the present retrieval systems primarily use two techniques including classic key-based full text search and SQL-like class retrieval. The former is not able to promise the precision. And the latter could not satisfy users since it requires their in-depth understanding of the system.In this paper, we present an information retrieval model for ontology-based EKMS in the framework of our ontology based configurable knowledge management system. We study the enterprise semantic networks and use the spread activation algorithm to retrieve the instances related to the query in the enterprise ontology base. Furthermore, We also analyze the query expansion approaches commonly used at the present times. Then to design the query expansion method based on the domain ontology, this paper proposes two different ontology concept categorizers including the one based on training documents and the another based on enterprise domain ontology structure. In the end, we implement an experiment prototype system associated with the model proposed before. The results of the following experiments verify the model's usability and show the improvements over the traditional information retrieval based on the key words.