基于搜索结果的个性化推荐系统研究

搜索引擎返回的信息太多且不能根据用户的兴趣提供检索结果,使得用户使用搜索引擎难以用简便的方式找到感兴趣的文档。个性化推荐是一种旨在减轻用户在信息检索方面负担的有效方法。文中把内容过滤技术和文档聚类技术相结合,实现了一个基于搜索结果的个性化推荐系统,以聚类的方法自动组织搜索结果,主动推荐用户感兴趣的文档。通过建立用户概率兴趣模型,对搜索结果STC聚类的基础上进行内容过滤。实验表明,概率模型比矢量空间模型更好地表达了用户的兴趣和变化。

A Study of Personalization Recommendation System Based on Search Result

It is difficult for the users to find interested documents in a simple and effective way by using search engines,since the returned information from search engines is too large and the engines are commonly hard to provide the users with required results based on their interests.Personalization recommendation is a valid method for lightening the user's burden on information retrieval.The paper presents a personalization recommendation system based on search result by combining content-based filtering technology and document-clustering technology,trying to recommend interested documents on their own initiative for the users by organizing search results automatically with the clustering method.By the founding of probability interest model of the users,the system realizes the content-based filtering of search results based on STC clustering.Experiments indicate that probability model can outperform VSM in expressing interests and changes of the users.