Scientific Computing and Data Management

Scientific computing and data management research examines how complex, multi-step computational processes—called workflows—are designed, executed, tracked, and shared across research communities. At its core, the work addresses a practical crisis: as science increasingly depends on large-scale data analysis pipelines, results are often difficult to reproduce because the exact steps, software versions, and intermediate data transformations go unrecorded. Researchers in this area develop systems for capturing data provenance (the full history of how a result was produced), building interoperable infrastructure, and exposing computational services through semantic standards that allow disparate tools to communicate. Open questions center on how to make provenance capture lightweight enough for routine use, and how to design workflow management systems that remain flexible across domains as different as genomics, climate modeling, and particle physics.

Works

466,253

Total citations

569,967

Keywords

Scientific WorkflowsReproducibilityData ProvenanceWorkflow ManagementBioinformaticsSemantic Web Services