OWL-T: A Task Ontology Language for Automatic Service Composition

In order to satisfy incremental business demands, it is often required to combine functionalities of several services together. A number of approaches for service composition have been therefore proposed in both academic and industrial communities. These approaches, such as BPEL, WSCI, etc. can be categorized into static, manual service composition methods. In addition, by applying Semantic Web technologies, many research works have been investigated to support automatic service composition. Despite of the significant results being achieved, the task of service composition is still a challenging and complex issue. The main reason is that the former approaches require too much detail and technical interventions for defining business processes while the latter approaches are not much scalable for sophisticated applications. In this paper, we will introduce our approach for developing an ontology/language based on the OWL, called OWL-T (T stands for task), which can be used for users describing and specifying formally and semantically their needs at a high-level abstraction, which can be then transformed into executable business processes by underlying systems. The OWL-T aims at facilitating the modeling of complex demands or systems without regarding details of low-level and technical aspects of underlying infrastructure.