A Method That Mines User’s Search Intent And Recommends Related Queries Base On Search Engine Logs

With the rapid development of the Internet, users need to face more and more data, and to find needed information from them, the only way is to use the search engine. However, most users face to tens of thousands of results returned, often do not know where to navigate, as returned results are always too long, too complex and many of them are not match user’s search intent. At the same time, the traditional search engine linear list results certainly increases the user’s burden to find needed information and reduces the user’s query efficiency. Nowadays improving search efficiency of the users is paid much attention. Researchers have proposed a variety of methods based on returned documents or search engine logs.The main research and exploration of this paper is improving the user’s query efficiency to Make up for the deficiency of the existing system. The method takes search logs into account, extracts needed information and get the relevant data set. Then construct seed query to extract candidate related words that meet the different intentions, and get features for training the classify model. For a new query, firstly get the related queries from the relevant data set with the classifier, and then get the final results with the short text similarity calculation method. Finally return more reasonable structure documents with related queries. On the basis of experimental analysis, we prove that the method can extract expected related queries and improve the efficiency.