View article

Today recommenders are commonly used with various purposes, especially dealing with ecommerce and information filtering tools. Content-based recommenders rely on the concept of similarity between the bought/searched/visited item and all the items stored in a repository. It is a common belief that the user is interested in what is similar to what she has already bought/searched/visited. We believe that there are some contexts in which this assumption is wrong: it is the case of acquiring unsearched but still useful items or pieces of information. This is called serendipity. Our purpose is to stimulate users and facilitate these serendipitous encounters to happen.

This chapter presents the design and implementation of a hybrid recommender system that joins a content-based approach and serendipitous heuristics in order to mitigate the overspecialization problem with surprising suggestions. The chapter is organized as follows: Section 2 presents background and motivation; Section 3 introduces the serendipity issue for information seeking; Section 4 covers strategies to provide serendipitous recommendations; Section 5 provides a description of our recommender system and how it discovers potentially serendipitous items in addition to content-based suggested ones; Section 6 provides the description of the experimental session carried out to evaluate the proposed ideas; finally, Section 7 draws conclusions and provides directions for future work.