AI and graph-based systems.
Managing data and software provenance for complex and dynamic simulations is a central aspect of eScience. Provenance represents the artifacts associated with a given set of simulations, including executed software, software version history, inter-relationships among difference software versions, simulation parameters, as well as input and output data. The changing behavior of a simulation due to changes in its code poses a risk to output data validity as software bugs, or other changes, can result in erroneous output. This problem is compounded by the fact that erroneous output may be used for the development of hypotheses that lead to future simulations. Providing data and software provenance enables scientists to not only determine precise experimental conditions of previous simulations, thereby increasing confidence in results, but also to detect when outputs may be erroneous due to bugs
Graphitti Workbench is a Scientific Workflow Management System that tracks the connections among experimental artifacts — including both software and data — helping researchers to understand the interactions among mathematical models, algorithms, software implementation, simulation configuration/parameters, and simulation results. The net effect is greater confidence in the quality of one's results by increasing the visibility of changes that may render the results invalid.
While provenance is a common component of scientific workflow management tools, they generally do not provide graphical user interfaces (GUIs) that use provenance to guide workflow. Graphitti Workbench was built to center workflow around its PROV-DM based ProVis software and data provenance visualizer, making provenance more readily accessible to scientists. This provenance graph provides improved visualization and the ability to replicate and derive activities from provenance. This enables the investigator to derive new experiments via simple interactions. Using ProVis as the system desktop highlights the importance of provenance to inform future experiments and improves provenance understandability. Making provenance the focus of scientific workflow implementation enables the diverse requirements of eScience workflows to be satisfied, as they have a common basis in provenance.