The Personalized Qurey Refinment Based Concept Lattice

With the rapid development of Internet, the information on it increases in exponent. Intelligent information retrieval and personalized service system become the most important research points on information service recently. How to design the reasonable and intelligent information retrieval system is the base of applying intelligent information retrieval and personalized service system.Since the document-term relation used in information retrieval can naturally be seen as a formal context of formal analysis (FCA), some research about studying information retrieval by concept lattice had been performed since 80's. Nowadays, with the recent advance in the theory and practice of concept lattice, concept lattice has become even more appealing and practical for information retrieval.The paper firstly does research about the problems of key-word search technology and inaccurate search results in search engine. The paper concludes the needs of intelligent information retrieval: interaction and concept expression. After analyzing the user modeling technology, this paper points out the problems of the automatic user modeling technology and proposes an algorithm of concept lattice user modeling. Then making use of the advantages of concept lattice of concept retrieval and visualization, this paper proposes a method of query expansion in search engine based concept lattice. Last, the paper provides the implementation of the system of query expansion based concept lattice.The main contribution of the paper includes the following:(1) Doing research about concept lattice and three different algorithms of building concept lattice. Providing the implementation and the analysis of the algorithms. (2) Proposed algorithm of concept lattice user modeling, which can solve the problems of the automatic user modeling technology. The advantages of concept lattice based user-modeling algorithm are presented: concept expression, visibility and easy manual extent and maintains.(3) A new query expansion system based concept lattice is designed and implemented in this paper. Making use of the concept retrieval and visibility structure, the new query expansion in search engine is conceptive and intelligent. Users of the search engine go through in the concept lattice whose first concept node is built by initial query and extend or refine the query through interactive mode. Users can dig the relation among the concept nodes and query node through the visible structure. This paper proposed functions, algorithm, flows, processes, interface design and result of the query expansion system based concept lattice.The new method of query expansion proposed by this paper can recommend new queries and interests to the users. The method is up the development of the search engine and information retrieval.