Association rule in data mining and an application on hepatitis

This study amed to develop a decision support system for physicians in diagnosing hepatitis by testing the association relationships of the disease and symptoms with Apriori algorithm. Totally 17 symptoms of 155 hepatit patients were analyzed. Application of Apriori algorithm to hepatitis data set, association relations have been established among some of the symptoms of hepatitis patients. The confidence levels of the association rules was observed between 95% and 100%. After analyzing, the probability of coexisting symptoms were calculated. These rules may provide statistically decision-making support to physicians in the diagnosis of hepatitis, which may have different prognosis and symptoms.