Improving Healthcare: A Data Mining Approach

Paul Becker, University of Oregon


The evolution and development of information systems over the past half century supported decision-making in various fields. Those fields range from biology to management. However, decision support is considered to be a difficult task in healthcare because those processes are complex and data relationships are hard to model. In this study, we identify the major difficulties when developing forecasting models to support decision-making and suggest an application in healthcare. We describe a model that relates data to information to knowledge using various terms and concepts. We follow a taxonomy that facilitates the generation of knowledge to improve decision making. Specifically, we produce knowledge related to diabetes. Diabetes is considered one of the most frequent diseases in the United States. Identifying diabetic patients is therefore very important. Our main goal is to create a core application that helps identifying diabetes.