Research On B2C Oriented Web Documents Clustering System

Rapid expanding of B2C e-commerce also brings a series of problems. For example, it has been very difficult for users to search what they want to get from the search engines of e-commerce website or the common search engines because of information exploding. Bad user experience must be disadvantageous for B2C developing. The method of Web documents clustering, which is a main branch of web mining, may be an excellent solution. This paper proposes an attempt of using Web documents clustering method on B2C e-commerce and a simplified application structure of Web documents clustering system, whose kernel function have implemented in programming. This paper is organized as follows.Firstly, this paper proposes two combining application approach which are pages recommendation and goods recommendation. Pages recommendation can be used in search engines of B2C websites and common search engines.Secondly, after analyzing two application structures of Web mining system, we propose the simplified application structure of Web documents clustering system. Detail processes are also given.Thirdly, we analyze three main technology, which includes Web documents collecting, text representation and text clustering process, in implementation process of Web documents clustering system. The meta-search engine technology and Web crawler technology are used in Web documents collecting; the Vector Space Model is used in text representation; the hierarchical clustering methods are use in text clustering process. Other clustering method can also be easily combined in the system. From an experiment, which is base on upwards technology, some practical advices are also given.At the end of the paper, we summarize our work and point out some future research work.