Should I Stay or Should I Go?: Exploiting Visitor Movements to Derive Individualized Recommendations in Museums

Nowadays, mobile recommender systems running on user's smart devices have become popular. However, most implemented mechanisms require continuous user interaction to provide personalized recommendations, and thus weaken the usability. This paper provides an innovative approach for taking advantage of user's movement data as implicit user feedback for deriving recommendations. By means of a real-world museum scenario a beacon infrastructure for tracking sojourn times is presented. Then we show how sojourn times can be integrated in both collaborative filtering and content-based mechanism approaches. An exhaustive experimental evaluation shows the suitability of our approach.