IMPACT OF SERENDIPITY AND DIVERSITY ON RECOMMENDATIONS

Abstract

Recommender systems are prominent tool for digging the online information to discover contents that are of value to the user. However, emphasizing only on accuracy to asses such systems leads to the problem of filter bubble in the recommendations. Therefore, inclusion of diversity and serendipity is vital for a great user experience. This review paper explored state-of-the-art research on the impact of both diversity and serendipity on recommender systems, examining how these concepts are defined and their balance with accuracy to enhance user satisfaction and broaden their horizons. This study aims to provide concise literature to inform researcher an understanding of not only considering accuracy but also to balance it with both diversity and serendipity in order to increase user engagement.