New Evolutionary Approach to Business Process Model Optimization Aliasghar Ahmadikatouli, Majid Aboutalebi Abstract-In today competitive business world, organizations and enterprises need to manipulate their business processes. The real key to be successful in these organizations lies in proper business process design and management. Concentrating on business process optimization and improvement, enterprises can achieve reduced costs, increased quality of products, raised efficiency of products, adapting with requirement changes, and they will flourish in this competitive environment. Since more than one objective is involved in business process optimization, multi objective optimization can be used appropriately. In this paper we proposed an evolutionary approach to optimizing business process model. Proper effective operators to generate new models are suggested as well. Index Terms-Business process (BP), BP optimization, BP modeling, multi-objective optimization, Genetic algorithm I. Introduction oubtless, business process plays an important role in enterprise progress. Hence, process modeling is one of the most essential steps in advanced enterprises. In addition, generating application and information systems are strongly dependable on the business process modeling. Many attempts have been made on creating variety of process modeling techniques, different notations, methods and tools each of which views process modeling in particular way and contains its specific semantic concepts. There exist different modeling approaches that undertake different aspects of a business process. Among those, few methods are able to analyze quantitatively and optimize a business process [5]. Modeling techniques can be divided in three different groups, mathematical models, business process languages and graphical languages that have been elaborated completely in [5]. On the basis of being graphical model and having been supported by a strong mathematical background, petri net is apt to be better option to be optimized. Business process analyzing has no value if it cannot help to improve or to optimize a business process. An ideal approach toward business process is, capturing a business process and providing appropriate tools to identify bottlenecks and to evaluate the performance and finally generate optimized business process based on specific objectives. However, the last part is usually overlooked if not completely disregarded. We are convinced to use efficiently this view, multi objective optimization, in the area of business process as it has been researched and analyzed in variety of computer science NP-Hard problems and it shows promising results. The rest of the paper is organized as follows. Section two provides some related works and methods. Manuscript received Jan 12, 2011; revised Feb 06, 2011. Aliasghar Ahmadikatouli, Islamic Azad University, Sari branch, Sari, Iran (e-mail: ahmadikatouli@gmail.com ). Majid Aboutalebi, Islamic Azad University-Science and Research, Tehran, Iran (corresponding author; e-mail: aboutalebi@iausari.ac.ir ). Section three describes some key concepts and contains business process modeling with petri net. Section four presents a new approach in business process optimization by genetic algorithm. Finally, section five concludes the paper. II. Literature Review There are different definitions of business process in business process literature that each one regards only a part of business process that correlated with analyzing, evaluating or modeling. Therefore, there is no commonly agreed definition. Some of definitions are too general that cannot consider all aspects of a business process [2] such as Harvey’s definition which proposed in [3] “step by step rules specific to the resolution of a business problem”. Another definition exists from Hammer et al [17] - “a business process is a set of processes that receive one or more inputs and generate a valuable output for customers,” and recently in [18] a new definition has been described with this implication, a set of activities and resources if has been sequenced properly, can do a business transaction. Other definitions can be found in [10, 19, 20] as well. According to Volkner and Warners in [4], since business process modeling organizes a process and analyzes current and alternative activities comprehensively and systematically, business process modeling is indispensible. Zho and Chen in [8] proposed that business process optimization leads to reduced process completeness time and running costs, as well as increasing quality of products and customer satisfaction. With this outlook, quite literally an organization can acquire the competitiveness advantage which it was looking for. Based on Moon and Seo in [9] the most attractive property of evolutionary algorithms is its flexibility in utilizing different objective functions with less mathematical requirements. In addition, Vetschera and Hofacker [6] have put in some effort to optimize a buesinss process with genetic algorithm, but as their method mainly depends on various mathematical formulas and has required a great deal restrictions, feasible solutions were hardly produced. However, Tiwari et al [13] and Vergdis et al [11] expanded this mathematical model and proposed a multi objective optimization algorithm that has been reported satisfactory results which opened promising researches as future works. Afterwards, Tian et al in [12] suggests four types of evaluation criteria including execution time, cost, throughput and queue length by analyzing and examining different optimizing parameters based on static and dynamic configuration. Valiris claimed that most of business process reengineering methods lack a formal confirmation that makes us ensure the generated model is the most suitable one for business process [7]. Therefore, need to a systematic approach that software redesign can have an appropriate model with following some steps, convinced researchers to propose new methods to find optimized business process D Proceedings of the International MultiConference of Engineers and Computer Scientists 2011 Vol II, IMECS 2011, March 16 - 18, 2011, Hong Kong ISBN: 978-988-19251-2-1 ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online) IMECS 2011