Improving Tag Search Based on User Activeness Scores
Young-Seok Lim, Kang‐Pyo Lee, Hyun-Woo Kim, Jae-Min Ahn, Hyoung-Joo Kim · 2011
With the advent of Web 2.0 technologies, social tagging service has gained a great popularity. The tags used in tagging systems are a simple list of keywords describing the resources on the web. Even though their structure is simple, tags abstract documents well and are regarded as useful metadata in the field of information retrieval, since they are produced cooperatively by many users. Specially, to improve the performance of search, a number of approaches using tags have been active to date. Depending on the participation of the users to assign the tags, the quality of tags as metadata varies remarkably. Furthermore more active and less active users exist. Because of this, the activeness of each user assigning the tags will also affect the search results. In this paper, we consider the degree of user participation in order to improve the search performance. We observe the del.icio.us social tagging service, from the perspective of the users activeness. And we propose a modeling for a variety of individual user activities. In addition, we propose an algorithm improving tag search by scoring the user activeness. Finally, we evaluate the proposed method to verify improving the satisfaction of the actual search through experiments.