Enhancement, Deployment, and Generalization of Metadata Technologies

Steve Harris ( Oxford University )

The Software Engineering Group at Oxford are developing a model- and metadata-driven architecture for research informatics. The architecture is being evaluated for use in large-scale clinical trials on both sides of the Atlantic, and is being integrated with the NCI cancer Biomedical Informatics Grid (caBIG); it is being enhanced and generalized for a wider range of applications, and for use in other scientific disciplines. We present the achievements to date, and the lessons learnt in developing frameworks for semantics-driven data acquisition and processing. We report on the deployment of the (open) architecture on widely-used commercial technology - Microsoft Sharepoint, Office, and .NET services - aimed at the expectations and requirements of a wide range of stakeholders. We discuss extensions of the approach to the model-driven development of laboratory information systems, including semantic annotation of tissue collections and microarray data. We discuss techniques for identifying candidate semantic metadata elements from existing artifacts (developed in collaboration with the Veteran's Health Administration) and the deployment of federated metadata registries (developed in collaboration with caBIG). We explain how the approach may be generalized to cover semantics-driven data acquisition and processing in other disciplines. (Joint work with Jim Davies, Charlie Crichton, Peter Maccallum, Lorna Morris)