Purpose: This paper deals with user-generated Interest indicators (e.g., ratings, bookmarks, tags). We answer two research questions: can search strategies based on Social Information Retrieval (SIR) make the discovery of learning resources more efficient for users, and can Community browsing help users discover more cross-boundary resources. By crossboundary we mean that the user and resource come from different countries and that the language of the resource is different from that of the user's mother tongue.Design: We focus on a portal that accesses a federation of multilingual learning resource repositories. A measure for user's efficiency in finding relevant resources was defined. We then collected users' attention metadata and use this empirical data to answer two hypotheses.Findings: We show that users are more efficient with Social Information Retrieval strategies, however, Community browsing alone does not help users discover a wider variety of cross-boundary resources.Practical implications: By social tagging and bookmarking resources from a variety of repositories, users create underlying connections between resources that otherwise do not cross-reference. This is important for bringing them under the umbrella of SIR methods. Future studies should include testing SIR methods to leverage these user-made connections between resources that originate from a number of countries and are in a variety of languages.Originality: The use of attention metadata to model the ecology of social search adds value to the actors of learning object economy, e.g. educational institutions, digital libraries and their managers, content providers, policy makers, educators and learners.