Proposed Architectural Model for Optimal Transformation of Decision Table and Decision Tree into Knowledge Base

Knowledge is one of the most precious resources of an organization. Every organization wishes to preserve and fully utilize its knowledge. Within organization knowledge is present in various forms, may be in the minds of workers or in documented form. In documented form the knowledge has various representation schemes such as frames, scripts, lists, decision trees and decision tables etc. We have proposed a transformation method to optimally transform knowledge present in decision trees and decision tables into knowledge base. According to our proposal, the knowledge present in these two representation schemes should first be converted into corresponding set of human interpretable rules by using some existing transformation algorithms. The decision tree should be converted directly into set of human interpretable rules but for decision table transformation, two ways are adopted; either it should be converted first into decision tree and then to set of rules or, should be converted directly into set of rules. Once the set of rules is obtained then it will be optimized by using some existing optimization algorithms and unnecessary conditions will be omitted. After rules optimization process, comparison with the existing rules in the knowledge base is carried out. If these optimized rules are not found in the knowledge base, it should be added. If some rules need updation, then these should be added after updation. During comparison those rules which already exist in the knowledge base should be omitted.