Reliable Retrieval of Top-k Tags

摘要: Collaborative tagging systems, such as Flickr and Del.icio.us, allow users to provide keyword labels, or tags, for various Internet resources (e.g., photos, songs, bookmarks). These which a rich source of information, have been used in important applications resource searching, webpage clustering, etc. However, tags are provided by casual users, so their quality cannot be guaranteed. In this paper, we examine question: given r set user-provided associated with r, can correctly described the k most frequent tags? To answer question, develop metric top-k sliding average similarity (top-k SAS) measures reliability tags. One threshold is then estimate whether sufficient retrieving Our experiments on real datasets show that threshold-based evaluation SAS effective efficient determine considered high-quality r.