An intelligent approach of knowledge searching within Internet-based distributive knowledge integrated environment

Used to integrate all kinds of knowledge coming from various locations, domains and disciplines, a framework of the Internet-based distributive knowledge integrated system (DKIS) is proposed. Since knowledge repositories are the key components of the DKIS, the efficiency and reliability of the knowledge searching affects the running performance of the DKIS system. We stress methods of efficiently retrieving knowledge from DKIS repositories. Based on artificial neural networks, fuzzy logic and rule reasoning, a novel approach to knowledge searching over the Internet is presented. The issue of knowledge layering is also discussed. Using the rolling bearing as an example, a prototype of knowledge searching is realized. The main aim of this study is to identify an efficient and reliable method of knowledge retrieval for the DKIS within an Internet environment so as to support its perfect operation. © 2003 Elsevier Science Ltd. All rights reserved.