Four-Fold Table Calculi for Discovery Science

Abstract

The goal of KDD is to search for interesting and useful patterns in data. Some of these patterns concern relations of two Boolean attributes ϕ and ψ. Such relations are evaluated on the basis of a four-fold table (FFT for short) 〈a, b, c, d〉 of ϕ and ψ. Here a is the number of objects satisfying both ϕ and ψ, b is the number of objects satisfying ϕ and not satisfying ψ, c is the number of objects not satisfying. and satisfying ψ and d is the number of objects not satisfying neither ϕ nor ψ. An example of such pattern is an association rule [1]. Four-Fold Table Predicate Calculi (FFTPC for short)will be introduced. Formulae of FFTPC correspond to patterns evaluated on the basis of FFT. FFTPC are defined and studied in connection with development of GUHA procedures, see e.g. [2,3,4]