Desing And Implementation Of Query Expansion Module In Search Engine

With the surge in network information, the users’requirements of the recall and precision ratio continue to increase. Base on the community information platform, we design and implement the query extension module to help user to find information which satisfy his needs in huge amount of dataaccurately. It will extend the users’query before searching, and lead to an improvement of the comprehensiveness and accuracy of retrieval results. In addition, this paper use the user’s interests to optimize the search results, in order to meet the personalized requirements. The main work summarized as follow:(1)We extends the words entered by user by using the query expansion module. Calculate the words similarity base on the HowNet, and find the synonymous of the query word as the candidate expansion terms. Search the Baidu Index of every candidate expansion terms for further screening, so it can effectively avoid the unrelated word added to the expand focus.(2) Extract the user’s history folders and favorite folders, then extract content, data cleaning, etc., receive the messages contain the user interests. Base on the similarity betweenthe documentvectors, we can do the text clustering among the document sets. The results of clustering can be used to represent the classifications of the user’s interest characteristics. The query expand module and personalized module studied in this paper have been integrated into the search engine, lead to the effective improvement of the search engine, and also enhance the user experience and user viscosity.