IntersectionExplorer, a multi-perspective approach for exploring recommendations

Abstract

In this paper, we advent a novel approach to foster exploration of recommendations: Inter-sectionExplorer, 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 Intersec-tionExplorer 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 In-tersectionExplorer as an effective tool, and indicate that its multi-perspective approach to exploring recommendations has great promise as a way of addressing the complex humanrecommender system interaction problem.

Generating and Understanding Personalized Explanations in Hybrid Recommender Systems

“…IntersectionExplorer [10] proposed a new tool that allowed for users to explore the provided recommendations in a multi-perspective way (personal, social, and content relevance). The proposed tool mixed recommendations from four different recommender systems (tag based, bookmark based, external bookmark based, bibliography based) along with data generated by humans 31:7 (e.g., tags, bookmarks) and provided the users with a visualization that allowed for exploration.…”

“…In the future, we plan to support an interactive recommender system that will use the observations from interactions with the user and adjust the explanations based on the user's preferences. In addition to that and inspired by IntersectionExplorer [10], we plan to implement a similar scalable visualization that allows users to explore recommendations as well as explanations.…”

Service-Aware Personalized Item Recommendation

“…Venn diagrams, scatter plots, UpSet matrix, text, ... [21], [22], [23], [24], [25] Review-based recommenders Item aspects Yes No presentation [26], [27], [28], [29], [30], [31] Review-based recommenders Rating conversion Yes No presentation [32] Review-based recommenders Item aspects Yes…”

A Service-oriented Perspective on the Summarization of Recommendations: Preliminary Experiment

“…MyMovieFinder [20] separately shows the recommenders that support a suggested item, while RelevanceTuner [38] uses stackable bars to integrate this type of information into a compact view. TalkExplorer [39] and IntersectionExplorer [4] use bidimensional graphs, or grid layouts, to show multiple dimensions of relevance. Moreover, Venn diagrams are used to overview suggestions [16] and they are combined with color bars to identify the recommender systems which contribute to the suggestions [30].…”