Recommendation Explorer was a recommendation system I worked on with Miles Efron when we were doctoral students at the University of North Carolina at Chapel Hill. It used knowledge discovery techniques to improve its representation of item-item relationships (Miles' contribution), and provided a graphical user interface to enable users to explore recommendations in the context of their information needs (my contribution).
The goal of Recommendation Explorer was to minimize user effort while producing high quality, personalized recommendations. Our approach was two-pronged: First, we used knowledge discovery techniques to develop an effective recommendation engine that required minimal user input, utilizing dimensionality reduction techniques to discover high-order relationships between resources, thereby producing quality recommendations based on sparse data. Second, we developed an interface that enabled the user to explore, manipulate, and preview recommended resources. Stored, modifiable profiles and a small number of simple widgets enabled the user to provide input to the system quickly and easily and thus personalize the recommendations to his or her immediate context.
Some of the interaction design ideas I implemented in the Recommendation Explorer project, such as mouseover previews of movie metadata, informed development of the AgileViews framework I focused on in my doctoral dissertation, and later became common in web-based interfaces, such as Netflix's.
