Time-Aware Recommender Systems: A Systematic Mapping

A Recommender System (RS) provides personalized suggestions of objects of users’ interest or that they may like. Traditional RS techniques consider only aspects related to users and items to recommend and ignore contextual information. Context-Aware RS (CARS) consider information about the user’s context to improve the recommendation process. Time is adimension of context that has the advantage of being easy to collect, since almost any system can record the interaction timestamp. Moreover, time can serve as valuable input for improving recommendation quality. Therefore, this work aims to investigate how time is being applied in CARS and, for this purpose, we used a Systematic Mapping methodology. In total, 88 papers were considered to answer the research questions defined. Initially we observed that the papers’ distribution by year have been increased in the last years. As a result, we also defined seven categories of how CARS uses the time in recommendation process.