Evaluating Term Weighting Schemes for Content-based Tag Recommendation in Social Tagging Systems

Social tagging systems allow users to publish different type of resources, such as Web pages or pictures, annotate them using keywords or tags and share their resources with other users. These systems achieved widespread success on the Web on account of the simplicity for organizing resources using open-ended tags. Recently, tag recommendation strategies have been proposed to alleviate the problems of ambiguity, syntactic variations and noise in tags cause by the inherent characteristics of natural language. In this work we proposed a content-based approach that generates a list of suggested tags for annotating a given resource starting from an analysis of its textual content exclusively. Thus, the proposed method can be used in situations in which there is not enough information for creating a tag-based user profile or compare the user with others. For extracting the more relevant words different term weighting approaches were evaluated, particularly considering the HTML structure of Web pages and the grammatical category of words in order to determine promising tag candidates. Experimental results of applying this technique to tag recommendation using several term weighting approaches are reported and compared.