A Scalable Tag-Based Recommender System for New Users of the Social Web

Abstract

Folksonomies have become a powerful tool to describe, discover, search, and navigate online resources (e.g., pictures, videos, blogs) on the Social Web. Unlike taxonomies and ontologies, which overimpose a hierarchical categorisation of content, folksonomies empower end users, by enabling them to freely create and choose the categories (in this case, tags) that best describe a piece of information. However, the freedom afforded to users comes at a cost: as tags are informally defined and ungoverned, the retrieval of information becomes more challenging. In this paper, we propose Clustered Social Ranking (CSR), a novel search and recommendation technique specifically developed to support new users of Web 2.0 websites finding content of interest. The observation underpinning CSR is that the vast majority of content on Web 2.0 websites is created by a small proportion of users (leaders), while the others (followers) mainly browse such content. CSR first identifies who the leaders are; it then clusters them into communities with shared interests, based on their tagging activity. Users’ queries (be them searches or recommendations) are then directed to the community of leaders who can best answer them. Our evaluation, conducted on the CiteULike dataset, demonstrates that CSR achieves an accuracy that is comparable to the best state-of-the-art techniques, but at a much smaller computational cost, thus affording it better scalability in these fast growing settings.