A new approach to the design of knowledge base using XCLS clustering

Abstract

A Knowledge Base is a special kind of data base used for storage and retrieval of knowledge. From the perspective of knowledge creators, maintenance and creation of knowledge base is a crucial activity in the life cycle of knowledge management. This paper presents a novel approach to the creation of knowledge base. The main focus of our approach is to extract the knowledge from unstructured web documents and create a knowledge base. Preprocessing techniques such as tokenizing, stemming are performed on the unstructured input web documents. Meanwhile, Similarity and redundancy computation is performed for duplicate knowledge removal. The extracted knowledge is organized and converted to XML documents. XCLS clustering is made on XML documents. Finally, Knowledge base is designed for storing extracted XML documents. A query interface has been developed to retrieve the search knowledge. To test the usefulness and ease of use of our prototype, we used the Technology Acceptance Model (TAM) to evaluate the system. Results are promising.