Design And Implementation Of Personal Search Engine

With the rapid development of Internet technology, the network can provide more andmore information to people. Search engine has been widely used in many fields, which istreated as a tool that people can get information on World Wide Web. In order to providepersonalized search service for users, personalized search engine comes forth. Because ofmany researchers contribution, people have made great progress in personalized searchengine. This thesis points out the shortage of current search engine and user’srequirements of personalized search, does some research on search engine and itstechnology, and designs a personalized search engine model that is based on user’sinterests mining. The main tasks of the thesis are as follows:Firstly, this thesis researches and implements the personal model. It analyzes thecharacters of personalized search, researches the search engine and its technology, anddesigns a model that is based on user’s interests mining. The model gets user’s interestsfrom the pages that he/she has visited before, classifies the pages according to the sameinterest, and manages the user’s interests according to the interest types. As to the storageof user’s interests, this paper consults the ODP catalog structure,establishes user interesttree to store the user’s interests, and uses the short interest and long interest to describe theuser’s interest characters. In order to reflect user’s interests changing in time, this paperupdates user’s interests using the algorithm based on for getting mechanism.Secondly, this thesis researches and implements other modules of personalized searchengine besides the personal model. The modules include Chinese segmenting, searchextending, network spider, index establishment and update and result ranking. This thesistakes the implementing difficulty of related technology and user’s requirements ofpersonalized search into account, and brings forward the algorithms of the modules above.At last, experiments are conducted to verity the efficacy of the personalized searchengine designed above. The contents of the experiment include two parts: theestablishment of personal model and the personalized search.