Some challenges for context-aware recommender systems

Recently, context-aware recommender systems (CARS), which incorporates contextual information into recommender systems, has become one of the hottest topics in the domain of recommender systems. In this paper, we identify and discuss some challenges for context-aware recommender systems, including viewing it as a process, valid contexts discovering and computing, contextual user preference elicitation, classification and design of context-aware recommendation algorithms, lack of publicly available datasets, evaluation, the sparsity problem, taking account of interdisciplinary research and applications. If these issues can be properly addressed, the development of context-aware recommender systems, in our opinion, will be promoted significantly.