Analysis Of Personalized Search Engine Based On User Interest

The huge amounts of data which is brought in by the development of Internet,provides a great deal of information for users,however,it also adds more difficulties for users to retrieve useful information.Because traditional information retrieval tools(search engines) which did not take users' interests into consideration,therefore the search results often were unable to meet the specific needs of different users.In order to solve the limitation of the traditional search engine, this paper propose a personalized search engine based on users' interests,this model expresses and manages users' interests by mining users' interests information,and through constant updates and maintenance,optimizes model gradually in order to enable them to accurately reflect the user's interests and needs,and provides basic analysis for the next personalized search.Major work includes the following:1.Dive into the architecture of search engine,analyze the key technology of building search engine and realization mechanism comprehensively.Design a highly configurable,flexible and generic framework which can be used in middle website, and finish the most functions.2.Dive into the web log mining technologies,discuss how to use the technology in user interest mining,and finally analyze how to apply the results of users' interests mining to achieve the performance optimization and search service personalization.3.This paper uses the outstanding open-source framework Nutch and Lucene as the development base,and introduces the construction and optimization of search engine modules in details by citing some examples.4.Finally we make function and performance testing the system,and the experiment result shows the system's performance is so good that it is suitable for being applied into medium-sized web site search.This article mainly contains information retrieval theoretical research and practical experience in projects;and they are valuable references for the development of information retrieval,data mining and user interesting construction technology research.The personalized search engine has wide range of application value and academic value.It can be used as search engine for some small and medium sites.