Context-Aware Multi-criteria Recommendation Based on Spectral Graph Partitioning

International audience; Both multi-criteria recommendation and context-aware recommendation are well addressed in previous research but separately in most of existing work. In this paper, we aim to contribute to the under-explored research problem which consists in tailoring the multi-criteria rating predictions to users involved in specific contexts. We investigate the application of simultaneous clustering based on the application of a spectral partitioning graph method over situational contexts in the one hand and criteria in the other hand. Besides, we conjecture that even with similar criteria-related ratings, the importance of criteria might differ among users. This idea leads us to use prioritized aggregation operators as means of multi-criteria rating aggregations. Our experimental results on a real-world dataset show the effectiveness of our approach.