Automatic preference learning on numeric and multi-valued categorical attributes

One of the most challenging tasks in the development of recommender systems is the design of techniques that can infer the preferences of users through the observation of their actions. Those preferences are essential to obtain a satisfactory accuracy in the recommendations. Preference learning is especially difficult when attributes of different kinds (numeric or linguistic) intervene in the problem, and even more when they take multiple possible values. This paper presents an approach to learn user preferences over numeric and multi-valued linguistic attributes through the analysis of the user selections. The learning algorithm has been tested with real data on restaurants, showing a very good performance. Nowadays it is practically unconceivable to select our summer holiday destination or to choose which film to see in the cinema this weekend without consulting specialized sources of information in which, in some way or another, our preferences can be specified to aid the system to recommend us the best choices. That is because we live in an era where there are so many data easily available that it is impossible to manually filter every piece of information and evaluate it accurately. Recommender Systems (RS) have been designed to do this time-consuming task for us and, by feeding them with information about our interests, they are capable enough to tell us the best alternatives for us in a personal-ized way. The preferences of the user are stored in a structure called user profile. In this work, as usual in the literature, it will be considered that each decision alternative is represented through a set of values assigned to a certain set of predetermined attributes or criteria. In these situations, the user profile must somehow represent the preference of the user with respect to each of the possible values of the attributes. With this information, the RS may rate and rank the corpus of available decision alternatives and show it to the user to help him/her to make the final choice. The representation of the preferences, the recommendation process and the automatic management of the dynamic evolution of the preferences are three of the most challenging issues in the development of this type of systems [24]. Concerning the latter problem (how to learn automatically the preferences of the users), the RS requires some kind of information from the users to guide the learning process. This feedback may be obtained implicitly, explicitly or combining both approaches. …