Developing the Search History Sharing System for Collaborative Information Retrieval based on Analysis of Information Search Process

我们分析了用户搜索历史的使用情况，关注用户的信息搜索过程。结果，考虑时间序列可视地表示搜索历史并且通过树结构示出搜索历史是有效的。此外，我们建立了一个共享的历史系统的书籍，可以代表用户的喜好，使用Nippon Decorative分类。作为实验的结果，证明了我们的系统能够以比诸如向量空间方法的传统技术更充分的准确度推荐对用户有用的人。

We analyzed use of the users' search history, paying attention to users' information search process. As a result, it was effective that the search history is expressed visually in consideration of a time series and is shown by a tree structure. Moreover, we build a shared history system for the books that can represent user preference using Nippon Decimal Classification. As a result of the experiment, it turned out that our system can recommend persons useful for the user with accuracy more sufficient than the conventional techniques, such as a vector space method.