Generating explanations in recommender systems based on matrix factorization techniques using context

Abstract

Users face difficulties to choose products and services on the Web because of the wide range of choices. In this context, the recommendation systems aim to help users identify items of interest in a set of options. Traditional approaches focus on recommending more relevant items to individual users, not taking into account the users' context. However, in many real-world applications, using contextual information is also important. For that, context-aware recommendation systems are used, since there are several studies that indicate that the use of such contextual information can improve recommendations. Among the traditional and context-aware recommendation systems, there are those that use matrix factorization techiniques. These techniques have become popular because they combine good scalability with good accuracy, and offer the flexibility to model various real-world situations. However, they are models of machine learning that do not yet offer transparency in the recommendation process, making it difficult for users to trust in the presented recommendations. In this sense, providing explanations for the generated recommendations tends to increase the user's confidence and satisfaction with respect to the system. So, this project aims to propose methods that use contextual informations to create explanations for the recommendations generated by recommendation systems based on matrix factorization. It is expected that the generated explanations increase user's satisfaction. (AU)