Modeling business process of web services with an Extended STRIPS Operations to detection feature interaction problems runtime

Abstract — Service-Oriented Computing achieves its full potential when services interoperate. Current service-oriented computing research is concerned with the low level interoperation among services, such as service discovery, service composition etc. However, a high level research issue in form of the feature interaction problem is challenging the interoperation of services. Traditional feature interaction methods are focused on the service design phase using formal methods or pragmatic software engineering analysis. Autonomy and distribution of service development and deployment create needs for runtime detection and resolution of feature interactions in SOC. This paper investigates the detection of feature interactions in web services at runtime and proposes ESTRIPs, an extended STRIPS operation to ensure conflict-free services are identified to populate business processes, using a combination of OWL-S, SWRL and runtime data extracted from SOAP messages. First, we define the feature interaction problem in business process during their execution and then introduce the ESTRIPS method. The implementation of a prototype is illustrated and a real world scenario shows the plausibility of our method for detecting feature interactions in business processes