Generating Synthetic Data for Context-Aware Recommender Systems

Context-Aware Recommender Systems (CARS) have emerged as a different way of providing more precise and interesting recommendations through the use of data about the context in which consumers buy goods and/or services. CARS consider not only the ratings given to items by consumers (users), but also the context attributes related to these ratings. Several algorithms and methods have been proposed in the literature in order to deal with context-aware ratings. Although there are lots of proposals and approaches working for this kind of recommendation, adequate and public datasets containing user´s context-aware ratings about items are limited, and usually, even these are not large enough to evaluate the proposed CARS very well. One solution for this issue is to crawl this kind of data from e-commerce websites. However, it could be very time-expensive and also complicated due to problems regarding legal rights and privacy. In addition, crawled data from e-commerce websites may not be enough for a complete evaluation, being unable to simulate all possible users´ behaviors and characteristics. In this article, we propose a methodology to generate a synthetic dataset for context-aware recommender systems, enabling researchers and developers to create their own dataset according to the characteristics in which they want to evaluate their algorithms and methods. Our methodology enables researchers to define the user´s behavior of giving ratings based on the Probability Distribution Function (PDF) associated to their profiles.