RuDES: A Semantic Method for Rules Dependency Extraction

Abstract

Nowadays, a variety of domains take advantage of rule based systems. In practice, such systems are continuously evolving by acquiring new knowledge and including more rules. The automation of the rule bases management has therefore become a must. Such a task becomes more fluent when based on the analysis of dependencies between the rules. This research work, introduces three main novelties while tackling the issue of rules dependencies detection. First, it amends the existing work. Second, it presents a new technique based on following the knowledge flow defined by the ontology. Third, it propounds a new way to represent the rules dependencies. To check the correctness and performance of our work, a prototype is developed and applied on three different ontologies from different fields.