Diversity in recommendation system: A cluster based approach

Abstract

The recommendation system is used to process a large amount of data to recommend new item to users, which are achieved using the many developed algorithms. Hence, it is a challenging task for lots of online applications to establish an efficient algorithm for a recommendation system that follows a good trade-off between accuracy and diversity. Diversity in recommendation systems is used to avoid the overfitting problem as well as excellent skill, which provides a recommendation based on increasing the quality of user experiences. In this paper, we proposed a methodology of recommendation to the user with diversity. The impact of diversity on the system leads to user experience for new items. The aim of this paper is to provide a brief overview of diversification with state of the art. A further similarity measure based on heuristic similarity measure “proximity impact popularity” is used to provide a new model with the better-personalized recommendation. The proposed approach gives profitability to many applications for better user experience and diverse item recommendations. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021.