Theapplication Of Particle Swarm Optimization In Query Optimization

With the rapid development and popularity of Internet, web has become an important way to find information. Since the information resources on the Internet increase rapidly, an important research focuses on how to effectively deal with these great capacities of information and acquire data that we need. By using information retrieval, information which is related with users'requrements is gained from information aggregation. Due to the ability that users are familiar with the system and describe query is different, results which are obtained when users use traditional technology, are difficult to meet users'requirements. This paper mainly conducts research on information retrieval query optimization based on Particle Swarm Optimization(PSO).Firstly, the dissertation introduces the background of the research, the current research status of information retrieval and query optimization, describes the content of the research and structure of the paper. Then the common Swarm Intelligence algorithm is introduced and the principle and application of Particle Swarm Optimization is detailed, in addition, the feasibility of PSO applied in the query optimization is analyzed. That is the basic theory of the research.Secondly, the application of Genetic Algorithm(GA) in the query optimization is studied and the advantage of PSO is analyzed. After that, the paper describes three classical retrieval models in information retrieval technology and probes to make PSO applied in query optimization. By using results which are got by initial query, PSO is used to allocate weight of keywords in the quey. Optimal query is got.Thirdly, the basic principle of relevant feedback is introduced and the relevant feedback technology based on the vector space model is learned. An example is found that accuracy of retrieval results is improved by introducing the relevant feedback to query optimization. By analyzing the traditional relevance feedback technique's effect on query optimization and skillfully making PSO applied in relevant feedback technology, the dissertation propose a new relevant feedback technology based on PSO, and it is applied in query optimization.Finally, with two indicators including the recall and precision to measure the search results, the query optimization methods proposed by the dissertation are validated. Experimental data indicates that ability of system is improved.