Font Size: a A A

Research On Pivotal Technology Of Focused Search Engine

Posted on:2007-09-29Degree:MasterType:Thesis
Country:ChinaCandidate:Z Q YangFull Text:PDF
GTID:2178360212495421Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the rapid development of Internet, the WWW has become an indispensable enormous information space to exchange information. In the face of such tremendous flood of information, people often lost themselves in the required information. How to find out the information they need fast and accurately has become a depressing problem.With the drawback of existing search engines, a solution to the topic search engines is proposed, and it satisfies what the search engines' specialization required. As the series of problems about theory and technology mentioned in the solution, follow research has been done:Firstly, a model framework of the topic search engine is improved and operated principles are given. Based on the achievement of thematic search engines, the special topic service of the search engine is realized.Secondly, text automatic classification technology in this paper is an important part in developing special topic search engine. In allusion to the shortage in text automatic classification, more illuminations are given to the improvement and perfection of the feature extraction technology, features weighted technology, importation word extraction technology and the log analytic technology, which ensure the improvement in completion and precision of the designed topic search engine.Thirdly, classification is an important research aspect of the topic search engine. A practical classification method based on data view was adopted for search engine. In the mean time, the participle dictionary of special subject has been constructed, which offers convenient searching for users, and the working efficiency has been raised.Finally, after analyzing the deficiencies of the traditional k-average clustering method, a text clustering algorithm is put forward. It can better improve the text clustering by selecting the better initial clustering center. It is proved to improve the stability and the targeting mechanism results by selecting the better initial clustering center.
Keywords/Search Tags:Information clustering, Focused search Engines, Chinese word segmentation, Vector space model, Special dictionary
PDF Full Text Request
Related items