Business Artifacts Discovery and Modeling Zakaria Maamar 1 , Youakim Badr 2 , and Nanjangud C. Narendra 3 1 Zayed University, Dubai, U.A.E. 2 INSA-Lyon, Lyon, France 3 IBM Research India, Bangalore, India Abstract. Changes in business conditions have forced enterprises to continuously re-engineer their business processes. Traditional business process modeling approaches, being activity-centric, have proven to be inadequate for handling this re-engineering. Recent research has focused on developing data-centric business process modeling approaches based on (business) artifacts. However, formal approaches for deriving artifacts out of business requirements currently do not exist. This paper describes a method for artifact discovery and modeling. The method is illustrated with an example in the purchase order domain. Keywords: Artifact, Data, Discovery, Process, Operation. 1 Introduction Continuous changes in market opportunities and conditions have led enterprises to re-engineer their business processes. Typically, these business processes are modeled in an activity-centric manner. While this way of modeling is popular, it has several limitations, e.g., aligning business requirements to business processes is not simple and modifying these processes mid-stream is cumbersome. A data- centric approach through (business) artifacts [3] can address these limitations. As in our earlier work [2], we adopt the definition of artifact from [3] as a “concrete, identifiable, self-describing chunk of information that can be used by a business person to actually run a business ”. That is to say that an artifact is a self-describing collection of closely related data that represent a business record, which describes details of goods and services provided or used by the business. One would consider order and menu as artifacts when modeling a restaurant. An artifact is subject to changes that are reflected on a state transition system called Artifact Life-C ycle (ALC ). Transitions between successive states in an ALC are the result of executing specific tasks in a business process. There is an abundant literature on artifacts [1,3,4]. However, there is still a lack of rigorous approaches that assist those in charge of discovering and model- ing artifacts We propose a method that examines the discovery of artifacts from three perspectives. The data perspective capitalizes on the data in a system and the dependencies between these data. The operation perspective capitalizes on the operations in a system and the dependencies between these operations. Out of the data and operation perspectives, two separate lists of candidate business P.P. Maglio et al. (Eds.): ICSOC 2010, LNCS 6470, pp. 542–550, 2010. c  Springer-Verlag Berlin Heidelberg 2010