基于文本分类与用户兴趣的个性化搜索与推荐的研究与实现-计算机软件与理论专业论文.docxVIP

Research Research and Implementation of Personalized Search and Recommendation based on Text Categorization and User Interest Computer Software and Theory Name：Cheng Ming Supervisor：Professor Tang Yong ABSTRACT Web has become one of the main approaches for US to obtain information．But it is difficult for US to get the information we need from the Web due to its fast growing and its characteristic of great magnitude，dynamic，unstructured and out·of-order．So how tO be user ccntr／c and provide a method for users to get the information they really need and the information that is authoritative is an important problem faced by the researchers． In this dissertation，the technologies of Search Engine，Authoritative Web Search, Text Categorization,User Interest Modeling，Recommendation based Oil Cooperative Filtering and some related technologies，such as Chinese Word Segmentation， Webpage Parsing and Information Filtering,are carefully researched and integrated- This dissertation suggests a solution of the Personalized Search and Recommendation based on Text Categorization and User Interest and implement its prototype called Alpha． In this dissertation，the Authoritative Web Search algorithm HITS is research and improved in Some way．A corpus for Text Categorization of some scale is constructed and a classifier is trained using the corpus．A user interest model based on VSM and its real time update algorithm are designed and SO is the translation between users’ long-term interest and short·term interest．A method is provided for personalized search using users’information acquirement habit and user interest which is described extensively and multi—granularity．Also personalized recommendation is provided using using technology of cooperative filtering． In the prototype Alpha,the techniques of multi—thread and data caching ar。“8ed to shonen the time Alpha takes to response to the useIS．Cookie is used for the users’ convenience and real．time interest update is used to improve users