Medical Insurance Data Mining Using SPAM Algorithm

Abstract

The sequential pattern data mining technology is widely applied to various fields, and it brings an indispensable value for many areas, especially in the field of medical treatment. But the amount of health-care data is large and the information included is extensive, so some valuable information may have not been found, which needs us to take the further research. By using the Sequential Pattern Mining (SPAM) algorithm to deal with the health-care data, we try to find the user’s medical behavior and the doctor diagnosis model or rule. This article first introduces the characteristics and the worth of Medicare data and data mining on it, especially the sequence pattern mining significance. Then discusses the ideas and characteristics of SPAM algorithm and the advantages of high efficiency, we use SPAM algorithm to deal with health care data, and try to find the regularity of visiting doctor, medical treatment characteristics and drug-use mode about insured person in a certain period of time. The experiments reveal the treatment mode and characteristics of the drug of pregnancy, which can provide guidance and reference for the diagnosis and treatment.