Two systems are currently being developed that will make use of both CBR and DM. These systems started as CBR systems, and will have DM-methods added for further usage of the data in the cases. The systems have not been fully developed, and are only briefly described here.
Case-Method is a methodology currently being developed at NEC for building and maintaining large-scale CBR-systems [21], [22]. They claim that collecting, sharing and mining experiences from cases is the best approach to improve knowledge level and skills in an organization. The DM methods that will be used are statistical analysis and rule- extraction. They will also do change and deviation analysis on the case data.
At the University of Rostock, they are developing a system which shall discover health risks, and forecast epidemics, based on earlier knowledge in cases [11]. They will use deviation analysis to discover changes in data collected from patient data from hospitals. Sequences of scenarios are used as cases, which are then retrieved, reused, revised and retained. A scenario consists of information on the current public health situation, such as new cases of illnesses, the load of the health service and contextual information such as weather and season. DM techniques will be used to discover ``conspicuous'' changes in the current scenario sequence, using a statistical method called the G-test. To examine the change, CBR techniques are used.