The Modeling and Analyzing Methods of Weighted Knowledge Network for Domain Knowledge Based on Keywords Clustering

ABSTRACT

Keywords clustering, as the basic method of domain knowledge analysis, has some problems such as difficult to understand the clustered tree diagram, scarce of further analysis methods, etc. The paper proposed a new approach to analyze domain knowledge based on keywords clustering. The proposed weighted knowledge model (WKN) is composed of two types of nodes (nodes of high-frequency keywords and nodes of clusters which come from keywords clustering and named as keywords nodes). Based on WKN, some new methods are suggested to analyze domain knowledge, such as main sub-fields analysis and representation, important sub-fields and hot spots of domain knowledge identification, research fronts analysis, etc., and all the analysis results can be illustrated as a sub-network of WKN. In the end, a case study was conducted to verify the feasibility and validity of the methods. Compared with the existing methods, the proposed methods seem more clearly, deeply and conveniently, and present new tools for researchers to study and utilize domain knowledge.

Keywords: