Contextual Modelling in Context-Aware Recommender Systems: A Generic Approach

Context-aware recommender systems (CARS) use context data to enhance their recommendation outcomes by providing more personalized recommendations. Context modelling is a basic procedure towards this direction since it models the contextual parameters to be used during the recommendation process. Most literature works however build domain specific contextual models that only represent information of a particular domain, excluding the possibility of model sharing and reuse among other CARS. In this paper we focus on this issue and study whether a more generic modelling approach can be applied for CARS. We discuss a possible solution and show through literature review on relevant systems that the proposed solution has not yet been applied. Next, we present a novel generic contextual modelling framework for CARS, discuss its advantages and evaluate it.