Comprehensive variability modeling and management for customizable process-based service compositions

Abstract

Variability in process-based service compositions needs to be explicitly modeled and managed in order to facilitate service/process customization and increase reuse in service/process development. While related work has been able to capture variability and variability dependencies within a composition, these approaches fail to capture variability dependencies between the composition and partner services. Consequently, these approaches cannot address the situation when a customizable composite service is orchestrated from partner services which themselves are customizable. In this article, we describe a feature-based approach that is able to effectively model variability within and across compositions. The approach is supported by a process development methodology that enables the systematic reuse and management of variability. A prototype system supporting extended BPMN 2.0 is used to demonstrate the feasibility of the approach.