Modelo de repositorio para gestionar reglas electorales generadas mediante técnicas de minería de datos

Abstract

Having information on electoral behaviors at hand is very important for the different actors involved in an electoral process, since these, the behaviors, play a relevant media role during political campaigns. In this sense, in this article, it is proposed a repository model or meta-model for storing models and electoral rules generated through data mining techniques from databases that contain results of recent electoral processes, particularly, of the federal elections held last year in Mexico for the presidential election and this year´s election in different states of the country to elect governors, municipal presidents and local deputies. The proposed meta-model includes a set of structures and a series of processes that allow the models and rules generated to be stored, based on ad-hoc phrases, so that this information can be retrieved through phrase aliases, emulating in a certain way the execution of queries in natural language. Thus, this work has double contribution, on the one hand, the electoral knowledge generated through data mining and, on the other, the metamodel for the management of this knowledge.