Play It Again, SAM -- Using Scientific Workflows to Drive the Generation of Semantic Annotations

A key issue in eScience concerns the management of data used, transformed and produced by experiments. Such data are highly heterogeneous and require a wide range of mechanisms to support their creation, documentation, mining, management and visualization. This paper tackles one of these mechanisms: annotation of scientific data. It discusses specific aspects of a framework that helps scientific research by providing support to create and manage annotations of geospatial data sources. The annotation process is driven by scientific workflows, and is thus repeatable and customizable. The annotations produced are linked to ontologies by SAM (Semantic Annotation Manager), a system we developed to ensure annotation consistency and management.