Research Of Web-Based Personalized Information Search System

With the rapid development of Internet, the information on it increases inexponent. How to search out one's most interested contents demands irrelevantto users decides on whether we can utilize the huge information resources. It hasbeen the focus question of searching information on the Web, and it is also theresearch goal of our article.This paper proposes the main issues of information retrieval including thesystem structure of information filtering, the characteristic and classification offiltering systems, the relation between retrieving and filtering, the commonmodel of filtering system, the evaluation index of the system performance andso on. It dwells on the information retrieval questions as a whole and points outthe existent problems in current information retrieval systems.Aiming at the existed problems in current IR systems and based on theexistent IF technology, the filtering algorithm was improved and an individuatedfiltering system which has the intelligent characteristic, subjective ability andextendibility was designed and realized in this article. A composite filteringapproach was adopted in the new system. The main research issues of thisarticle are as follows: retrieving and filtering, a composite filtering algorithm,Relative Recall and Relative Precision.At first, a model of personalized search engines was presented. Workingprinciple of the model was given. According meta-search engine, Personalizedand intelligent service of search engines through application of informationfiltering technology in search engine was realized.The filtering algorithm is the core of an individuated filtering system. Theindividuation was realized through the filtering approach. A composite filteringalgorithm was designed in this article.High Recall in information retrieval on internet brings the gigantic results.People still need to look for their requisite information again in the results.When search engine provided personal information retrieval on web, Recall andPrecision are not suitable anymore. So Relative Recall and Relative Precisionwere described in this article.