A self evolutionary rule-base

The rapid growth in domain knowledge is the main reason for the evolution of knowledgebase’s maintaining the domain knowledge. Rule-based Decision Support Systems (DSS) are the most effected systems with the growing knowledge. The experts need to continuously update the rule-base for the new knowledge. This manual and periodic updates in rule-base are time consuming and less useful. In this paper we propose a Self Evolutionary Rule-base algorithm for rule-bases of DSS to decrease the burden from experts and also provide updated knowledge on time. To achieve this objective, we develop a generic structure for rules storage that not only provide efficient manipulation of rules but a generic structure for storage of rules regardless of rules nature/format. The detail working of proposed Rule-base system for rules storage and manipulation is provided in this paper. For the proof of concept, we have implemented the Self Evolutionary Rule-base algorithm in Socially Interactive Clinical Decision Support System (SI-CDSS). The focus is on diabetes disease patients and the overall SI-CDSS is deployed in Microsoft Azure environment. In its implementation, Rough Set generated rules are used and the algorithm is executed on Rough Set generated rules.