ViPen: A Model Supporting Knowledge Provenance for Exploratory Service Composition

In current service composition efforts, the intermediate products such as data and related process fragments are neglected in many cases when the deviation occurs from predefined composition. The lack of adequate provenance support makes it not convenient when building the composition in an exploratory fashion. In this paper, we present ViPen model, and illustrate its operations to enhance the provenance for flexible deviation in service composition. Meanwhile, a partially ordered relation “derivation” is formed which guarantees the new derived processes can reuse previous knowledge effectively. A case study of one bioinformatics experiment is explained to verify our work.