Recommendations by Collaborative Browsing

This paper advocates for a more extensive definition of recommender systems, stressing in particular the importance of information browsing and discovery. A recommendation tool, called "seeking for inspiration" is here presented. It is aimed at users that are seeking for suggestions, do not have complete knowledge of the tourism jargon and can be intimidated by more traditional search tools based on queries. Our tool, in the first place computes a small number of travel offers and displays them on a graphical user interface that relies on pictures to illustrate the offers. The user can browse the offers and provide feedbacks/critiques about them. The system uses a diversity-based retrieval to incorporate the user feedbacks and to suggest some more offers that are supposed to be a better match to user’s preferences. We report the results of the empirical evaluation that shows the usefulness of such an approach especially for travelers coming from more distant countries and those with a more holistic learning style