Improving Automatic Data Structure Generation for e-Science Applications

The usage of ontologies to develop a semantically rich experiment models promises to be a key advantage of scientific applications over earlier alternatives. Whilst it is often recognized that information gathered for the ontology modeling process can describe naturally the scientific knowledge and can be used for interoperation among heterogeneous systems (by establishing a global schema, for instance), it may also be used to create data structures, including database schema and initial code signatures, containing metadata and semantics for their applications. The aspects involved in the translation of ontology models into suited metadata, however, can render in wasted efforts and useless schemas for scientific applications. This paper explores an approach to generate semiautomatically appropriate data structures for handling scientific information. Based on this approach, we developed a tool that let scientists to develop canonical models and automatically generate the related database schema. This tool supports a wide range of scientific use cases for complex models within the VL-e project. This project carries out concerted research along the complete e-Science technology chain, ranging from applications to networking, focusing on new methodologies and re-usable components.