A Document and User Matching Model via Document Keyword Analysis

Abstract

Since more and more transactions are carried out over the Internet, it is critical for the enterprises to systematically manage the huge amount of documents and efficiently explore the target users in e-commerce networks. Concerning the high variety of document contents and user behaviors in cyberspace, it is not appropriate for the modern organizations to exploit the document and user characteristics simply by human effort. This paper develops an integrated approach to automatically and consistently determine the document/user categories according to the document keywords and browse history. Three modules including Keyword-Category Correlation Analysis (KCCA), Document Classification (DC) and User Classification (UC) are proposed to systematically match the documents and users. At the stage of KCCA, tabulated correlation coefficients between keywords and categories are established based on the existing corpus or document repository. Using the correlation table, the relationship between the target documents/users and specified categories is explored via DC and UC algorithms. A web portal as well as a demonstration case is provided to evaluate the effectiveness and performance of the proposed approach. The DC, UC and match analysis models can be used in CRM and KM systems for intelligent customer services (e.g., knowledge classification) and can be applied in the publication of documents and information.

• Document Classification
• User Classification Keyword Extraction
• Document Management
• Knowledge Management