Provenance as a prerequisite for reproducibility and replicability in GIScience
Topics:
Keywords: Geographic information science (GIScience), reproducibility, replicability, provenance, ethics
Abstract Type: Paper Abstract
Authors:
Jason Alan Tullis, University of Arkansas
,
,
,
,
,
,
,
,
,
Abstract
Advances in geographic information science (GIScience)-specific reproducibility and replicability (R&R) are complicated by geographic, computational, versional, regulatory, socioeconomic, and other factors. One way forward is to prioritize our capacity to understand, access, and manage geographic information provenance where it simultaneously impacts R&R as well as other stakeholder interests including fitness for use, quality, expression and propagation of errors, privacy, intellectual property, export control, and cybersecurity. When conflicts between stakeholder interests arise, questions about ethical use of the associated provenance will inevitably follow. The capacity to understand and certify actual and intended use of specific granularities of such provenance, without conveying its contents, may be a prerequisite for broad stakeholder trust in R&R. Context and illustrations are drawn from ecosystem services, Mars data analysis, agricultural imaging, and the emergence of geospatial knowledge graphs.
Provenance as a prerequisite for reproducibility and replicability in GIScience
Category
Paper Abstract