Towards Case-Based Support for e-Science Workflow Generation by Mining Provenance
ECCBR '08 Proceedings of the 9th European conference on Advances in Case-Based Reasoning
| Hi-index | 0.00 |
As the Web services technology matures and becomes widely employed in business areas, scientific communities have also started to use Web services as a means of collaboration across multiple institutions. Among the applications of Web services, creating a workflow that orchestrates various Web services to conduct an experiment is one of the most important features in scientific communities. Although a number of workflow composition tools have been developed and widely used in scientific communities, the current tools have limitations, such as a lack of interoperability between different workflow tools, a lack of abstract workflow representation that is independent of underlying workflow languages while supporting various scientific workflow patterns, and insufficient support for the experimental nature of scientific workflows. In this dissertation, we develop a scientific workflow framework based on Web services specifications to address the interoperability issue while investigating whether various dynamic features of scientific workflows can be supported by the current Web services framework. We provide a workflow composition tool, XBaya, as the core of the proposed workflow framework. XBaya has a simple graphical interface for domain scientists to use and is interoperable with other Web services-aware tools because it employs Web Service Description Language (WSDL) for workflow component definition and uses Business Process Execution Language (BPEL) as the core workflow language. The proposed workflow framework addresses multiple needs of scientific workflows, such as various workflow patterns used in scientific workflows, monitoring and modification of running workflows, and dynamic binding of Web services to adapt workflow to dynamically changing distributed environments. The usage of the proposed workflow framework in a large meteorological project demonstrates how each need of scientific workflows is satisfied.