Research Data Curation in Visualization

Dimitar Garkov, Christoph Müller, Matthias Braun, Daniel Weiskopf, Falk Schreiber

View presentation:2022-10-17T19:15:00ZGMT-0600Change your timezone on the schedule page
2022-10-17T19:15:00Z
Exemplar figure, described by caption below
Mapping the interplay between publication, curation, and research repeatability. Each publication is characterized by its own research pipeline, where data is produced or reused. Produced data is not only any collected or generated data, but also data from the research activity itself. In the curation pipeline, produced data sets are curated, ideally as these emerge, and deposited in data repositories for sharing and long-term preservation. The producer and depositor roles can be filled by different persons. To facilitate research repeatability, i.e. reproduction and replication, a two-step review ensures quality and formal policies compliance before the data set is finally stored.

The live footage of the talk, including the Q&A, can be viewed on the session page, BELIV: Paper Session 2.

Abstract

Research data curation is the act of carefully preparing research data and artifacts for sharing and long-term preservation. Research data management is centrally implemented and formally defined in a data management plan to enable data curation. In tandem, data curation and management facilitate research repeatability. In contrast to other research fields, data curation and management in visualization are not yet part of the researcher's compendium. In this position paper, we discuss the unique challenges visualization faces and propose how data curation can be practically realized. We share eight lessons learned in managing data in two large research consortia, outline the larger curation workflow, and define the typical roles. We complement our lessons with minimum criteria for selecting a suitable data repository and five challenging scenarios that occur in practice. We conclude with a vision of how the visualization research community can pave the way for new curation standards.