Integration of Spatial User-Item Relations into Recommender Systems

Recommender systems aggregate information about users and items to be recommended to generate adequate recommendations. This paper proposes two approaches to include information about spatial relationships of users and items in order to improve the quality of recommendations. The two approaches are compared with non-spatial recommendation using two synthesized data sets and four different evaluation metrics. The results show how spatial information can improve recommender results. Furthermore each of the two approaches can perform better than the other, depending on what assumption of user behaviour is represented by the data set. Special ramp-up problems, occurring when including spatial information into recommender systems and different application areas for spatial recommender systems are discussed.