抄録
The Specific Health Checkup aimed at preventing metabolic syndrome was initiated in 2008 in Japan. Through this guidance, people considered at high risk of metabolic syndrome's progressing are expected to be made aware of their own problems in their daily lifestyle choices and to improve their daily life behaviors by themselves. To support this national endeavor, we have introduced the idea of data mining techniques by building a Bayesian network (BN) corresponding to the framework and the data configuration of the Specific Health Checkup. In this paper, we analyze data on 11,947 anonymized individuals, extracting major factors from questionnaire data related to lifestyle and consider the possibility of generalizing our BN by introducing these lifestyle questionnaire factors as nodes of the BN. This proposed method is found to provide good performance, and its usefulness is revealed by evaluating the level of change of the responses to the questionnaire. Introducing the lifestyle questionnaire factors nodes into the BN would be expected to make it a more useful risk assessment tool.
