Tagging activity has been recently identified as a potential source ofknowledge about personal interests, preferences, goals, and other attributesknown from user models. Tags themselves can be therefore used for findingpersonalized recommendations of items. In this paper, we present a tag-basedrecommender system which suggests similar Web pages based on the similarity oftheir tags from a Web 2.0 tagging application. The proposed approach extendsthe basic similarity calculus with external factors such as tag popularity, tagrepresentativeness and the affinity between user and tag. In order to study andevaluate the recommender system, we have conducted an experiment involving 38people from 12 countries using data from Del.icio.us, a social bookmarking websystem on which users can share their personal bookmarks.
展开▼
