Method for automatic inference of data attributes of service composition activities and fragments
Manuel Carro <firstname.lastname@example.org> Universidad Politécnica de Madrid (UPM)
The goal is to automatically infer finitely-describable, domain-specific characteristics (related to e.g. privacy level or quaity) of data used by service compositions activities, based on characteristics of data inputs and composition structure. The approach takes into account service orchestrations that involve complex control structures, such as branches, loops, and sub-workflows. Abstract interpretation-based sharing analysis is applied to Horn clause representations of service choreographies and the results are interpreted in terms of concept lattices (from Formal Concept Analysis) featuring data and activities. The inputs to the analysis and the outputs from it are presented as tables that associate the domain-specific characteristics (attributes) with input data items, activities in the composition, and the intermediate and resulting data items.
We use BPMN notation for defining service orchestration workflow, and Ciao Prolog Development Environment for running the sharing analysis.
There is currently no stand-alone demo available.
Ivanović, D., Carro, M., & Hermenegildo, M. (2011). Automated Attribute Inference in Complex Service Workflows Based on Sharing Analysis. In Proceedings of the 8th IEEE Conference on Services Computing SCC 2011. IEEE Press.
Ivanović, D., Carro, M., & Hermenegildo, M. (2010). Automatic Fragment Identification in Workflows Based on Sharing Analysis. In M. Weske, J. Yang, P. Maglio, & M. Fantinato (Eds.), Service-Oriented Computing @ ICSOC 2010. Lecture Notes in Computer Science. Springer Verlag.
Service Composition, Data & Information Related Quality
Relationship with Future Internet and Internet of Services
Service compositions are a powerful and flexible mechanism for implementing more complex, cross-organizational processes based on component services (often provided and managed by a third party), and can, in turn, be used. Analyzing characteristics of data in such complex service systems is deeply related to the issues of data security and privacy, resource usage, fragmentation, and dynamic distributed enactment of complex business processes.
Relationship with Cloud