Extraction of rules from a contingency table in the case of complex Big Data type systems

Abstract

This article describes a modeling approach based on data from a contingency table. We develop a step-by-step procedure based on the definition of a knowledge extractor using entropic concepts. The limit extraction threshold is chosen according to a multi-criteria analysis. Rules in symbolic form endowed with a truth coefficient and an efficiency coefficient are then deduced and constitute the model of the syste