Web query term logs from Internet search engines provide an unobtrusive source for understanding users' search interests. To make the subject categorization of Web query terms more efficient and adaptive, an automatic or computer-aided mechanism is needed. The purpose of this paper is therefore to propose a mechanism that carefully integrates human and machine efforts to explore Web users' search interests. The approach is designed to consist of a four-step process including the extraction of core terms, construction of subject taxonomy, automatic subject categorization of query terms, and observation of users' search interests. Some experiments have been conducted, and the preliminary results show that it is possible to categorize query terms into a pre-defined subject taxonomy which helps to explore users' search interests by online monitoring the distribution of subject categories and the composed query terms in real time. The research findings are proved valuable for enhancing Web IR (Information Retrieval) systems and Web content organization.
|頁（從 - 到）||372-381|
|期刊||Proceedings of the ASIST Annual Meeting|
|出版狀態||已發佈 - 2001 十二月 1|
ASJC Scopus subject areas
- Information Systems
- Library and Information Sciences