A Knowledge Component Analysis Model Based on Term Frequency and Correlation Analysis

Abstract

Knowledge extraction, storage, management and reuse are the typical issues for an enterprise to maintain its competitiveness in the knowledge economy. Implicit knowledge of the domain experts is usually extracted and represented in the form of documents. The valuable knowledge lying in the documents should be explored and reconstructed by human beings through reading. Regarding the knowledge recognition issue, this paper proposes a model for document component analysis. In the proposed approach, two algorithms, namely frequency repository construction algorithm and document fragmentation algorithm, are developed. By construction of term frequency and correlation repositories, a comparison basis for document fragmentation is established. A web-based platform as well as a real case (i.e., analysis of intellectual property (IP) documents of automobile component designs) is provided to demonstrate the feasibility and performance of the proposed model. Concerning the importance of document contents and structure, the document fragmentation model can be applied in the enterprise knowledge management systems for efficient knowledge representation and acquisition.