“Our belief is that a useful measure of interestingness should generate index values that are reasonably distributed throughout the range of possible values (such as in a SND)”
- Case A: four equally-represented levels, each with probability 0.25;
- Case B: one dominant level (p1 = 0.97) and three rare levels (p2 = p3 = p4 = 0.01);
- Case C: three equally-represented majority levels (p1 = p2 = p3 = 0.33) and one rare minority level (p4 = 0.01);
- Case D: four distinct probabilities (p1 = 0.05, p2 = 0.10, p3 = 0.25, p4 = 0.60).