A comprehensive survey on comparisons across contextual pre-filtering, contextual post-filtering and contextual modelling approaches

Abstract

The significance of including contextual information in the recommendation process has been recognized for some time. Recently, there has been growing interest in the area of recommender systems (RSs), specifically in context-aware recommender systems (CARS). Methods for generating context-aware recommendations were classified into pre-filtering, post-filtering, and contextual modelling approaches. While there exists a substantial amount of research on CARS, little attention has been paid to compare the three different contextualization paradigms. In this paper, we present the several novel approaches of the different variant of each of these three contextualization paradigms and present a comprehensive overview of the state-of-the-art comparisons across contextual pre-filtering, contextual post-filtering, and contextual modelling approaches. We then identify key challenges that need to be addressed by the current RS researchers. This will help the academicians in developing a deeper understanding of their trade-offs and practitioners can use it to choose the best option according to their market strategy.