An Approach to Configuration Management of Scientific Workflows

A scientific software ecosystem aims to integrate all stages of an experiment and its related workflows, in order to solve complex problems. In this vein, in order to assure the experiment proper execution, any modification that occurs must be propagated to the associated workflows, which must be maintained and evolved for the successful conduction of the research. One way to ensure this control is through configuration management using data provenance. In this work, the authors use data provenance concepts and models, together with ontologies to provide an architecture for the storage and query of scientific experiment information. Considering the architecture, a proof of concept was conducted using workflows extracted from the myExperiment repository. The results are presented along the paper.