A context modelling system and learning tool for context-aware recommender systems

Abstract

A critical contextual modelling issue in context-aware recommender systems research has to do with developing domain and application specific models that offer no reuse and sharing capabilities. Developers and researchers struggle to design their own models without any guidance, often resulting in overspecialized, inefficient and incomplete context models. On the basis of a prior work, we have developed an online context modelling system and learning tool that is able to guide and teach CARS developers and researchers through the process of CARS context modelling, on how to build their own context models in a way that offers sharing and reuse, while at the same time it teaches them important modern concepts derived by CARS research, significantly advancing as a result their knowledge in the field. © 2013 Springer-Verlag.