Toward a Context-Aware Serendipitous Recommendation System

Recommendation system development has been an important domain in the industrial and academic fields for the past two decades. Recently, the importance of developing a context-aware serendipitous recommendation system has emerged. As such, we investigate the latent features of items that may be recognized by the users of such a system. We assume that users will move from one item to another through the latent features reflected in the sequence of items. Our work specifically focuses on the process of predicting the sequential and changing taste of users. We show the existence of latent features by presenting a topic map and suggest a context-aware serendipitous recommendation system.