In this paper, we present a system to support medical doctors to derive hidden effects of the classes that are organized to improve lifestyle related disease. Recently in Japan, the number of such kind of classes is increasing rapidly, but it is not easy to know the type of participants for whom each class has particular effects. Our system does a data mining to find a group for which a class has particular effect by generating several combinations of conditions from the data of the classes. With our system, one can find groups of particular effect without much time and labor, and can also confirm the original data from which the effect is revealed. Our system can be used as a fundamental analysis tool to analyze the data of that kind of classes.
View full abstract