个性化搜索中用户兴趣挖掘技术方法研究计算机科学与技术专业论文.docx

哈尔滨工业大学工学硕士学位论文 哈尔滨工业大学工学硕士学位论文 - - II - Abstract To some extent,traditional information retrieval may meet user needs, but the same query is different search intent for different times and different backgrounds,different areas of the user;personalized search provide personalized service for different users, Personalized Search is an important area of research, how to effectively identify the user interest category preferences is a challenging research subject,it is the key technology for we can effectively identify the users interest preferences in personal services;the problems that need to solve for user interest preferences mining include to collect user information, how to collect and organize these data information、how to representation and store those information; We need to use reasonable technology and user model to anlalyze for geting the user preference baseing on collected data information.and we research the problem that users information and interest how adaptive tracking and learning. This paper designs the user preference model that learn the users interest algorithm, our method main features aer effective not only to indentify general user preference categories but also indentify user prefereece category that belong to different catetory query. This paper research the method of user preference mining. Our algorithm idea is that use a query string that user input and can show the users search intent, then we mining the user interest preference through the queries content, but sometimes the user enter a short query string or some query strings mean ambiguity,and then we have an idea that user will read relevant docment,the documents are classified through text classification techniques and identify the class that the documents are. To indentify the users interested preference,We map the query to the class through document relationship.then it is able to more clearly identify the user interest category preferences.the user model can effective