A recommender system that deals with items having an image as well as quantitative features

A big part of data around us is in image format and people use these images in many of their decisions. The popularity of an item, in many cases, depends highly on its visual quality. For instance, the shape of a car has a significant influence on the attitude of potential customers toward it. Recommender systems try to provide people with recommendations resulted from an automatic process which is aimed at giving the users a better experience working with system, and perhaps improve the system owner´s sales. As images are quite important in users´ decisions, in this paper we have proposed a method to take images into account when trying to give the user a recommendation, which despite its apparent advantages has not found a fair amount of attention so far.