Study on Developing Personalized Web Retrieval Support Systems

为了开发考虑用户意图的Web检索支持系统，我们进行了利用社会书签、ODP（Open Directory Project）、Twitter、博客和浏览历史提取用户个人资料以及从Wikipedia中提取相关词的研究。我们开发了几个Web推荐系统，用户之间的几个相似之处提取的配置文件的基础上。我们还研究了词之间的时间距离，提取相关的词的距离的基础上，并显示了有效的词的及时检索。

We conducted the research on extracting user' s profiles using social bookmarks, ODP(Open Directory Project), Twitter, blogs and browsing history, and related word extraction from Wikipedia in order to develop Web retrieval support systems considering his/her intent. We developed several Web recommender systems using several similarities between users based on the extracted profiles. We also examined a temporal distance between words, extracted related words based on the distance, and showed the effectiveness of the words for a timely retrieval.