IntersectionExplorer, a multi-perspective approach for exploring recommendations

Abstract

In this paper, we advent a novel approach to foster exploration of recommendations: IntersectionExplorer, a scalable visualization that interleaves the output of several recommender engines with human-generated data, such as user bookmarks and tags, as a basis to increase exploration and thereby enhance the potential to find relevant items. We evaluated the viability of IntersectionExplorer in the context of conference paper recommendation, through three user studies performed in different settings to understand the usefulness of the tool for diverse audiences and scenarios. We analyzed several dimensions of user experience and other, more objective, measures of performance. Results indicate that users found IntersectionExplorer to be a relatively fast and effortless tool to navigate through conference papers. Objective measures of performance linked to interaction showed that users were not only interested in exploring combinations of machine-produced recommendations with bookmarks of users and tags, but also that this “augmentation” actually resulted in increased likelihood of finding relevant papers in explorations. Overall, the findings suggest the viability of IntersectionExplorer as an effective tool, and indicate that its multi-perspective approach to exploring recommendations has great promise as a way of addressing the complex human-recommender system interaction problem.