Creación de un sistema recomendador de artículos del portal de la tienda de oportunidades Día.

Abstract

Abstract: This degree work consists of the implementation of a recommender system for cleaning items for the user in a virtual shopping store where products of opportunity are offered, said research proposes to know the best product recommendations for users who currently use the channel opportunities of the Multinational DIA S.A. For the development of the project, a recommendation algorithm based on collaborative filtering was used, with the use of the K-close neighbors technique, where the system allows selecting the article depending on the best scores made by the users who have rated some product. For the performance evaluation of the algorithm, quality metrics such as RMSE are used to evaluate the predictions, precision and recall to evaluate the recommendations. Finally, it is mentioned that the project in question is an educational research and in the future the company will decide its implementation if it takes place.