Web用户兴趣模型研究-软件工程专业论文.docxVIP

郑州大学硕个学位论文 Abstract Wi伽性le development of the Intemet，search engine hωbeωme one of the best channels for the us町 aωess to infonnation. Howev町 due ωdifferentωers of different backgrounds and int町ests，the need for information resources has also been diffi町ent. The resources of user needs are often a subset of the field on the internet web. However， di伽ent users got the same results. Personalized s町vice become the ∞ncern，and become the hot. Research and development of p时sonalized services have a larg町 由ωretical significance and practical valueωcustomers fiωt and efficient aωω information of needed. P町sonalized recommendation technology is an effective way to solve the problem. Based on the users interests and characteristics of information resourc倒， it collects and classifi刨出e information， and providωor recommends information for the users. Personaliz创 search and reωImmendation system provid创 p即sonalized s即vices in two importantωpects. Toda弘曲的的 two kinds of personalized recommendation systems， one is 伽e rule-based system，and another 扭曲e information filtering system. The la忧町 one can be divided into ωntent based filtering system and collaborative filt创ng system. The \浏阳 int时est model is impo阳时也r the personalized recomm侃侃，tion system.The main research work of this pap町 is 部 follows. 1.This p啊 analysed two kinds of typical us町 interest models: based on the ∞时ent of the page to access 山町 int町创t model，and based on 剧目browsing behavior of user interest mode1.叽le work is the basis of user interest modelω 由e furth町 research.Based on behavior of the user; we found the user int町est model based on a joint matrix of the Web logs. 2. Today，ωer interest model of res侃rch based on the Web is focus叫on the 也eme notation，阳的 bookmark notation， and keywords list notation. A伽.we analyz创 the general user interest model and explored the neural network， we found user int町创t model bωedon 也e artificial neural n就wo也 mKey words: Web log; us町 inter刷刷del; 阿印刷ized;