The Research Of Key Technology In Personalized Information Retrieval Based On Internet

With the development of Internet and Web technology, Internet has been agigantic space of information exchange. Facing the gigantic information, Peopleoften lose their direction when they look for their requisite information. Howquickly searching for their requisite information from gigantic informationresource has been a difficult problem of perplexing the user. So this paper putsforward personalized information retrieval and puts the emphasis on theresearch of the crucial technology concerning with the personalized informationretrieval system.Firstly, we survey the major techniques for information retrieval andfiltering methods. We provide an overview of the traditional ones full textscanning, inversion, signature files and clustering. We discuss attempts toinclude natural language processing and neural network.Secondly, this paper analyzes Vector Space Model. This paper puts forwardan improved searching model by adopting an advanced Vector Space Modelbased on weight combined with thematic words and individual informationdatabase.Thirdly, this paper presents the work of the word segmentation andanalyzes the ways and difficulties to combine them. This paper gives a methodof extracting words without dictionary. This method avoids the complexgrammar and rules, which needn't enormous support from dictionary andresolves the problems brought by the new words. So we conclude that thismethod has better exactness and is very pragmatic and powerful in practicaloperation.In the end, the ability of facilitating users to achieve useful information ismore and more important for informational retrieval system. Aiming at existentproblems, a personalized informational retrieval system is introduced. Thissystem is implemented using the agent techniques and the relevant feedbackapproach. This system generally can be used in the information retrieval inInternet and have high significance in theories and applications.