Information Retrieval Model Based On User's Category Intention

Information Retrieval (IR) causes an enormous impact on knowledge acquisition, scientific research and lifelong education. People's IR knowledge and skills have become an indispensable part of the whole knowledge structure. The study to improve the result of IR system is of great significance.IR solution is facing two major issues. One is queries' independence problem and another one is data sparse problem. In addition, How to embed more semantic factor into the IR system is also a hot issue in the study.This paper discusses how to get user's intention from the user's query and tries to improve the result of IR system based on the intention. We firstly give the definition of user's intention and build models for documents to be searched and user's queries basing on specific intention. We then discuss some methods to improve the IR result based on intention. Finally, we design some experiments to validate different methods proposed in this paper. The experimental results show a good improvement by combining intention with IR.The innovation of the paper is the different study perspective from prior research. 1. Mine user's intention from queries. 2. Define intention specific and set up intention set for specific intention. 3. Use new IR unit building the system.