Semantically Linking and Browsing Provenance Logs for E-science

Abstract

e-Science experiments are those performed using computer- based resources such as database searches, simulations or other applica- tions. Like their laboratory based counterparts, the data associated with an e-Science experiment are of reduced value if other scientists are not able to identify the origin, or provenance, of those data. Provenance is the term given to metadata about experiment processes, the derivation paths of data, and the sources and quality of experimental components, which includes the scientists themselves, related literature, etc. Conse- quently provenance metadata are valuable resources for e-Scientists to repeat experiments, track versions of data and experiment runs, verify experiment results, and as a source of experimental insight. One specific kind of in silico experiment is a workflow. In this paper we describe how we can assemble a SemanticWeb of workflow provenance logs that allows a bioinformatician to browse and navigate between experimental compo- nents by generating hyperlinks based on semantic annotations associated with them. By associating well-formalized semantics with workflow logs we take a step towards integration of process provenance information and improved knowledge discovery.