Abdelhak Mansoul, Baghdad Atmani and Sofia Benbelkacem, Oran University, Algeria

Many systems based on knowledge, especially expert systems for medical decision support have been developed. Only systems are based on production rules, and cannot learn and evolve only by updating them. In addition, taking into account several criteria induces an exorbitant number of rules to be injected into the system. It becomes difficult to translate medical knowledge or a support decision as a simple rule. Moreover, reasoning based on generic cases became classic and can even reduce the range of possible solutions. To remedy that, we propose an approach based on using a multi-criteria decision guided by a case-based reasoning (CBR) approach.

Decision support, Case-Based Reasoning, CBR, Multi-criteria decision support, diabetes diagnosis, JColibri