本文对个性化建模技术进行了研究,提出一个基于用户兴趣挖掘的个性化模型,并将其用于个性化元搜索引擎中,为个性化分析提供依据。该模型基于客户端,将用户的访问页面作为挖掘对象,不需要用户过多的参与,自动从用户的隐式反馈中推导出用户的兴趣。模型中将用户的兴趣分时分段,分层分类的进行记录、管理,并仿照人类记忆的遗忘规律对用户兴趣进行有选择、非平等的遗忘,通过不断的更新与优化模型,使模型能准确反映用户的兴趣特征。 In this paper, personalized modeling technology is studied, and a personalized model based on user interest mining is proposed, which is used in personalized meta-search engine to provide the basis for personalized analysis. The model is based on the client and takes the user’s visit page as the mining object, and does not need the user to participate too much, and automatically deduces the user’s interest from the user’s implicit feedback. In the model, the user’s interests are divided into time-sharing segments and classified and recorded, managed, and modeled on the forgetting laws of human memory, users’ interests are selectively and unequally forgotten. By continuously updating and optimizing the model, Accurately reflect the user’s interest characteristics.