MEANS OF VISUAL MODELING AND CODE GENERATION FOR FUZZY RULES

Nikita O. Dorodnykh, Alexander Yu. Yurin, Sergey A. Korshunov, Dmitriy Yu. Sopp, Dmitriy S. Shpachenko

Matrosov Institute for System Dynamics and Control Theory of Siberian Branch of Russian Academy of Sciences, CentrSib LLC

Logical and associative rules are still the most common way to represent expert knowledge and support of decision-making. The use of rules in the case of inaccurate or uncertain information requires the development of specialized linguistic (language) means and software, providing their visualization and generation of source codes. In this paper, we propose a FuzzyRVML - an extension of a Rule Visual Modeling Language, designed for modeling logical rules containing elements of fuzziness. FuzzyRVML supports the fuzzy data type, the concepts of linguistic variable, term, and certainty factor. Descriptions of the main elements and constructs of FuzzyRVML are presented, as well as an illustrative example containing code generation in FuzzyCLIPS. Personal Knowledge Base Designer was used to support the notation proposed, as well for its implementation and evaluation.

fuzziness, fuzzy knowledge base, fuzzy rules, logical rules, code generation, rvml, FuzzyCLIPS

Back