Research And Implementation Of Query Expansion System Based On Users’ Feedback

Information retrieval is one of the hottest research fields, which studies how to search information needed from large amount data source. But because of the inaccuracy of the query submitted by users, the retrieval results are often inaccurate, and can not satisfies the users’need. Basing on this, I research and implement a query expansion system based on users’feedback to deal with the problem in a new way.Query expansion technology is a supplement and extension of information retrieval technology. This paper describes the development, characteristics and key theories of query expansion technology, and then mainly discussed the current technology and typical systems. This paper makes the web pages and text files as the study object to implement a query expansion system based on users’feedback. Because of the use of summarization to help users judge the relevance of documents and extract query expansion data, auto summarization is another key point of this paper. Using summarization, the feedback information of users can be achieved directly, accurately and rapidly, which makes up user interest model. At the same time, for the purpose of describing users’interest synthetically and exactly, the user interest model is in form of multiple hierarchy architecture, to describe user interest model of different stage and different characteristic. Finally, according to user interest model, the system finishes the work of query expansion, including recalculating the weight of documents and extracting query expansion keywords.Experiment proves the higher efficiency of the system. And it considers both the registered and unregistered users, so it has good practicability.