429 I International Symposium Engineering Management And Competitiveness 2011 (EMC2011) June 24-25, 2011, Zrenjanin, Serbia THE ROUGH SETS THEORY BASED EXPERT SYSTEMS Vladimir Brtka, Ph.D* brtkav@gmail.com Eleonora Brtka, M.Sc brtka@sbb.rs Visnja Ognjenovic, M.Sc visnjaognjenovic@gmail.com University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Zrenjanin, Department of Computer Science, Republic of Serbia ABSTRACT The paper deals with the decision rules synthesis in the domain of expert systems and knowledge based systems. These systems incorporate expert knowledge which is often expressed in the If Then form. As it is very hard for experts to formally articulate their knowledge, automated decision rule composing algorithms have been used. The rule composing algorithms are often based on the rough sets theory. Originally, rules are composed from data table by equivalence relation, while in this paper we investigate the rules based on dominance relation. The main goal of this paper is to single out possible benefits and advantages of dominance relation based rules over equivalence relation based rules. Key words: expert systems, knowledge based systems, the rough sets theory, decision rules INTRODUCTION The history of the use of computers in engineering problems goes together with the developments in computer hardware and software technology. Engineer utilize principles of science and mathematics to develop new technologies which are then used to create products, structures, machines, processes or even entire systems. It is well known that different tasks in engineering problem solving require different computational tools. Inference from a set of facts, which simulate intelligent decision making, plays a important role in some problem-solving tasks, which involve creativity, while creativity implies the ability to produce novel solutions which are better than previous solutions. These computational tools should be able to use expert knowledge of the problem domain for decision making. The software tools that involve expert knowledge and inference mechanism are called expert systems (ES). ES deals with knowledge processing and complex decision-making problems. The “bottleneck” of the expert systems is a problem of formal articulation of the expert knowledge. It is very hard for the expert to formally express knowledge. Broader based systems which use different knowledge sources are called Knowledge Based Systems (KBS). This paper investigates the If Then form based decision rules, defined or expressed by experts. These rules are readable and easy to understand (Luger, Stubblefield, 1993, Jones, 2008). It is well known that rule synthesis is possible by usage of the rough sets theory. Originally, the rough sets theory based rule induction used equivalence relation, while in this paper we investigate dominance relation. The main goal of this paper is to point out the differences between equivalence relation and dominance relation based rules. THE ROUGH SETS THEORY The rough sets theory was developed in the early 1980s (Pawlak et al.). This new approach proved to be very useful for the data analysis in various domains. It is of importance to artificial intelligence (AI) and cognitive sciences in the domains of machine learning, knowledge acquisition, data mining, decision analysis, expert systems, decision support systems, pattern recognition and inductive reasoning. Many