Proposition of Modified Attributable Pure Confidence for Exploration of Meaningful Association Rules

Association rule mining is used most frequently in data mining techniques. It is the method to quantify the interesting relationship between a set of items in a large database, and has been applied in various fields like healthcare, insurance, education, and internet shopping mall. There are many interestingness measures as the criteria for evaluating association rules. Among them, confidence is the most frequently used, but it has the drawback that it can not determine the direction of the association. The attributably pure confidence were developed to compensate for this drawback. But it is difficult to interpret operationally in negative association because its range has negative infinite. In this paper, we proposed a modified attributably pure confidence to be able to interpret operationally as an objective interestingness measure. And then we investigated the conditions of interestingness measure and some useful properties, and compared some properties of and confidence and attributably pure confidence through a few experiments.