Abstract
The common issue for medical information systems are missing values. Generally, gaps are filled by statistically suggested values or rule-based methods. Another approach is to use the knowledge of information systems working under the same ontology. The medical incomplete system receives a query unable to answer, because of some unknown patient attributes. So, it has to communicate with other medical systems. The result of the collaboration is collective knowledgebase. In this paper, we propose a measure supporting choice of closest pair of systems. It determines the distance between the two systems. We use ERID algorithm to extract rules from incomplete, distributed information systems. Each constructed rule has confidence and support. They allowed to determine the distance between a pair of medical information systems. The proposed solution was verified on the basis of several “manipulated” medical information systems. Next, the solution was verified in systems with randomly selected data. The satisfying results were obtained and based on them, the proposed measure can be successfully used in medical systems to support the work of doctors and the treatment of patients.