How do context-aware artificial intelligence algorithms used in fitness recommender systems? A literature review and research agenda

Abstract

Recommender Systems (RS) help the user in the decision-making process when there is a problem of plenty or lack of information. The context-aware recommender systems (CARS) incorporate contextual information to improve the efficiency of RS. Prior works on CARS have compared algorithms and reported application domains. However, the practical challenges involved in CARS implementation with technological advances have not been adequately examined. We chose fitness as the application domain as there is increased user adoption, and the complexity is high. It involves the dynamic interplay of multi-dimensional contexts and technological advances in contextual attributes and incorporates machine learning algorithms in the process of context detection. We conduct textual analysis and conduct a systematic review to identify developments. The concept map demonstrates the components of contextuality, the methods adopted, and the role of AI techniques in amplifying contextuality in the algorithm. The paper identifies the challenges and presents the research opportunities.