MINERÍA DE DATOS BASADA EN LÓGICA DIFUSA PARA LA INTERPRETACIÓN DE CONSULTAS VAGAS DEPENDIENTES DEL CONTEXTO LINGÜÍSTICO DATA MINING USING FUZZY LOGIC FOR THE TRANSLATION OF VAGUE QUERIES DEPENDING ON THE LINGUISTIC CONTEXT

This paper presents a method in order to a flexible query-answering system can obtain, independently and dynamically, the semantics of imprecise words used as constraints in database queries by scanning the available data in the database and using fuzzy logic. The system's inference engine, according to the linguistic context defined by each query, chooses a fuzzy set model among the predefined theoretical models for different syntactic patterns with which can fit the text of a vague query and considering different levels of granularity in the fuzzy discrimination. It estimates the value of fuzzy set parameters by using a nonparametric method in the fuzzy discrimination, avoiding the intervention of experts. This achieves that the systems are not only adaptable, but reliable thanks to the validity of the responses.