DYNAMIC RADIUS FOR CONTEXT-AWARE RECOMMENDER SYSTEM

In the field of transportation and smart cities, context-aware recommendation is a trending subject that targets users' satisfaction and comfort. The subject has developed from e-commerce to cover different domains including path planning at urban areas. Researchers were competed in delivering the level of contribution in many areas. Search radius is considered a major pillar at any context-aware recommender framework, however, in many studies the search radius is not having a major contribution in the delivered systems. In this paper, a dynamic search radius algorithm is introduced as part of a context-aware recommender framework. Agglomerations and competition effects are used to enhance search radius results. Deep neural network is a major artificial intelligence method used in this research to tackle cold-start problem and to improve recommendation outcomes.