Beyond Accuracy: Evaluating Recommender Systems by Coverage and Serendipity

Beyond Accuracy: Evaluating Recommender Systems by Coverage and Serendipity

Mouzhi Ge, Carla Delgado-Battenfeld, Dietmar Jannach. 2010. (View Paper → )

Coverage and serendipity. Based on a literature review, we first discuss both measurement methods as well as the tradeoff between good coverage and serendipity. We then analyse the role of coverage and serendipity as indicators of recommendation quality, present novel ways of how they can be measured and discuss how to interpret the obtained measurements. Overall, we argue that our new ways of measuring these concepts reflect the quality impression perceived by the user in a better way than previous metrics thus leading to enhanced user satisfaction.

Traditionally recommender systems are evaluated on their predictive accuracy. Clearly looking beyond accuracy there are other measures that could be considered. It’s important to understand is your system is allowing for serendipity - if not at the recommender level then at least at the product level. Increasing serendipity is often an important goal of workplace products.