V-BPMI: A variability-oriented framework for web-based business processes modeling and implementation

Abstract

This paper advocates a new paradigm, V-BPMI, for flexible web-based business processes modeling and implementation. This paradigm emphasizes both variability and composability, which are very important in large and evolving organizations. Through process modeling and process operationalization, it proposes to establish a real engineering phase of the process variability. It is based on concepts such as process lines, process variants, process fragments and contextualization that support adaptation and relevancy qualities. V-BPMI is also based on reuse of web-based business process fragments. Thus, we consider here business processes as large compositions of sub-processes or process fragments. Composition is realized on-the-fly according to the context constraints where the process is deployed. This paradigm is developed in AP-HM (Assistance Publique -Hôpitaux de Marseille) hospitals throughout process of circulation of the drugs. In this context, our objective is to improve the efficiency of such processes and also to increase their security level. Indeed, in such large establishments/enterprises as AP-HM, a business process can be performed in several medical units. Our approach allows deploying a business process in an adapted way in each one of these medical units.

Adaptation dynamique de processus métier. Application au circuit du médicament à l’AP-HM

“…Le métamodèle des concepts de V-BPMI et l'architecture de l'approche sont détaillés dans (Angles et al, 2013).…”

Extending BPMN 2.0 Meta-models for Process Version Modelling

“…Several typologies for classifying variability capabilities of languages or notations have also been introduced (Schonenberg et al, 2008), (Nurcan, 2008), (Weber et al, 2008), (Andonoff et al, 2013), and three main types of variability are identified: variability by design (or flexibility) for handling foreseen changes in processes, variability by deviation for handling occasional unforeseen changes and where the differences with initial process are minimal, and finally, variability by evolution for handling unforeseen changes in processes, which require occasional or permanent modifications in their schemas. However, existing languages and notations introduced for modelling this variability (Kradofler and Grepper, 1999), (Lu and Shadiq, 2006), (Zhao and Liu, 2007), (Hallerbach et al, 2008), (Lu et al, 2009), (Hallerbach et al, 2010), (Chaâbane et al, 2011), (Angles et al, 2013) are incomplete as either they deal with only one type of variability and do not address these three types of variability in a coherent framework, or they mainly focus on the behavioural dimension of processes, or they are too specific with low degree chance to be used, or finally they are dedicated to specific domains.…”

Enhancing Variability Modeling in Process-Aware Information Systems through Change Patterns

“…At enactment time, it needs to be guaranteed that process variants are executed according to the configured process variant model [van der Aalst et al, 2010b;Hallerbach et al, 2009;Ayora et al, 2012a]. In addition, this phase covers configuration decisions that may only be made during enactment time (i.e., dynamic configurations) [Angles et al, 2013;Murguzur et al, 2014]. Even though configuration is partially performed at enactment time, soundness should be ensured at design time.…”

“…Concerning the three other artifacts, different techniques for defining them exist. For representing variable process fragments, for example, features models (from software product lines) Montero et al, 2008;Schnieders & Puhlmann, 2007;Czarnecki & Antkiewicz, 2005] or goal models Angles et al, 2013] can be used. In addition, variable process fragments may be defined based on a set of process model components , a variant list [Meerkamm & Jablonski, 2011], or a set of prespecified change operations Lu et al, 2009].…”