Modelo de contexto para recomendação de rotas baseada em interesse

Abstract

Context: Route advisers are often used by people, but these advisers do not consider contextual elements when recommending, ensuring only that the route of least distance or better road conditions. However, there is a lack of recommendations that consider users’ interests and the context to generate routes that can take users out of the routine, recommending routes that have places that are of interest to the user, guaranteeing a more comfortable route to the user. Objective: to develop a contextual model that dynamically considers user interests during a route recommendation. Therefore, with the contextual model it is possible to design a set of behaviors that, given a given context of the user, is able to infer preferences by one route or another. Method: was first identified in the literature, which user interests and contextual information are relevant to referral systems. Then the contextual model was constructed to represent the dynamic behavior of the system according to the interests of each user. Thus the behaviors produced were represented in rules based on the Drools tool. From this representation, it was possible to perform a proof of concept through the Drools tool, in order to exemplify the functionality of the behavioral models generated in this work. Results: the behavioral models were generated in order to be flexible, so that new contextual elements could be inserted in the future. Conclusions: Although there are some shortcomings, they can be supplied to improve the contextual models generated, because the more contextual elements are inserted in the context model, the more the system will be close to being a custom system.