Font Size: a A A

The Research And Design On Key Technologies For Personalized Meta-Search Engines

Posted on:2008-06-29Degree:MasterType:Thesis
Country:ChinaCandidate:H X WangFull Text:PDF
GTID:2178360215469538Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
In the era of information explosion, the fast and efficient information searching in Internet is an arduous task. Search engineers enhance us the ability of locating and gathering information. But the existing search engines still have lots of deficiency on searching efficiency, information maintenance, information repetition, load of network and websites .So it's an eagerly demand to research and develop a search engine with engine with powerful function, wide covering, friendly interface and accurate feedback.The meta-search engine is the exactly required, which is an integrated search tool, a web search engine based an meta-search technology aims at how to take use of and extend the capabilities of its sub-engines to satisfy the needs of engine uses. Ordinarily, the procedure of a meta-search Engine includes three steps: database select (sub-engines select), document select and result merge. Comparing with ordinary independent web engine ,meta-engine has high possibility to achieve high coverage, precision and recall .On the other hand ,the efficiency and performance of a meta-search engine depends much on its inner administration and management policies, if it adopts simple result-merging and sub-engine-selecting modules ,neglects the individual demands and the necessity of providing intelligent service, the users will also face much difficulties to find out the desired web results.Considering the limitation of the existing meta-search engines and users'personalizing quests, then based on user profiles, this paper propose several algorithms for personalized search. It's main research work includes:(1)The article analyzes the main arithmetic of the meta-search engine technology(2)Extracted the topics of user's interests and analyzed categories of user interests. These categories represent the user's search intention and provide the foundation for search engines scheduling and results filtrating.(3)A personalized scheduling strategy based on user profiles is designed to select most latently useful search engines for user query.(4)An algorithm of results merging based on user profiles is proposed.In the end, we design and implement an experiment system of personalized meta-search engine. And on it a variety of experiments have been carried out and analyzed. At the same times, we present future search work facing users'requirements for search engine.
Keywords/Search Tags:Meta-Search Engine, Personalizing, User Model, Strategy Of Dispatching, Results Merging
PDF Full Text Request
Related items