Attention-Based Information Composition for Multicontext-Aware Recommendation in Ubiquitous Computing

The recommender system that offers useful information to the users becomes more important in ubiquitous computing. Context-aware recommender systems are a core component in ubiquitous application. Because of the greater number of contexts, the total amount of information is larger in multicontext-aware environments. Multicontext therefore requires a more efficient recommendation method. We propose an attention-based information composition for multicontext-aware recommendation. We explain that the recommendation process would be comprised of three steps. Our proposed method is capable of recommending accurately and rapidly.