A Study On The Application Of The Techniques Of Data Mining In Personalized Information Retrieval System

With the popularization of Web application, not only the amount of information on the web grows dramatically but also the content of information on the web is updated frequently. The problem people face with today is no longer lack of information, but how to find pertinent information to satisfy their specific needs. Although to some extent traditional Information Retrieval (IR) technologies can satisfy these needs, they are suffering from low recall and precision. Due to stressing on universality, the information retrieval services provided by most commercial search engines cannot give satisfactory results for queries from customers with different background and intention. Being directed against this situation, the author researched on the Personalized Information Retrieval System Based on Data Mining.First, the thesis conducts in-depth analysis of the current status of the research and development trends of information searching engine systems, explores the feasibility of applying the techniques of data mining to personalized information retrieval systems, and proposes an architecture of personalized information retrieval system based on data mining. Then functions of the system and corresponding key techniques for implementing the system are analyzed and explored thoroughly.Finally, the author gives a mining algorithm based on class association rule of Apriori (algorithm). Simulation results indicates that the mining algorithm based on the association rule(s) of Apriori (algorithm) can dig out customers'personalized information, and the quality of the personalized information has close relationship with the degree of supporting.