Personalizing Diversity Versus Accuracy in Session-Based Recommender Systems

One of the most important concerns about recommender systems is filter bubble phenomenon. While try to personalize information, they tighten around users and deprive them a wide range content. To overcome this problem, one can diversify personalized recommendation list. A diversified list usually presents broader content user. Session-based are types recommenders in which only current session user available, therefore, should recommend next item given items session. diversifying conventional has been well assessed literature, it gained less attention session-based recommenders. Diversity accuracy have negative correlation, i.e., by improving other will be declined. In study, we propose diversity enhancing approaches based on sequential rule mining k-nearest neighbor methods. Finally, performance balancing approach that improves both these systems. We demonstrate proposed methods four music datasets.