ACTIVE COMPUTER-MEDIATED SHARING AND DISCOVERY OF SCIENTIFIC KNOWLEDGE THROUGH ONTOLOGIES AND LOGICAL INFERENCE

Abstract

Sharing scientific knowledge is central to the process of knowledge synthesis and transfer. In order to increase the effectiveness of dissemination and retrieval of scientific knowledge, a four level framework for knowledge sharing, discovery and integration has been proposed that leverages new information and computer technologies, including semantic reasoning engines, ontology engineering for knowledge representation, and computational modeling tools. By leveraging computational models widely used in science as medium for expressing knowledge, knowledge sharing and integration through computers and the Internet could be made dynamic and interactive. Semantic matchmaking based on ontologies and logical inference is used to share computational models and other knowledge resources in a form that can be interpreted semantically by computers to provide more effective support for searching and discovery. A Web-based knowledge sharing platform called EKOSS, for expert knowledge ontology-based semantic search, has been created to implement this semantic matchmaking through logical and rule-based inference using description logics ontologies. The computer intelligence resulting from logical inference could make possible an active form of computer-mediated sharing and discovery, which is illustrated through an analogy of a library. Finally, a method for using the EKOSS system to implement this active form of sharing and discovery of scientific knowledge is demonstrated.