MIC Framework: An Information-Theoretic Approach to Quantitative Association Rule Mining