A decentralized and service-based solution for data mediation: The case for data providing service compositions

SummaryThe last few years have seen an increase in the use of e-services as a natural consequence of the service-based economy growth. However, the distributed nature of services can raise several heterogeneity problems that can hamper the widespread adoption of service-oriented architectures (SOA). In this paper, we are interested in data heterogeneity problems that can occur when using composite services. We propose a service-based approach for automatically inserting appropriate mediation services in service compositions to resolve structural heterogeneities in their data flow. This mediation is ensured through a special kind of Web services called data mapping Web services. In addition, in order to be in agreement with the distributed nature of the Web, we propose a decentralized solution to publish/discover these services. We demonstrate how our approach applies to the data structural heterogeneity problem in data providing service compositions as a particular data heterogeneity case. Provided experimental results show that our approach is efficient in realistic situations.