Guerrilla Badges for Reproducible Geospatial Data Science (AGILE 2019 Short Paper)

Abstract

The building blocks of research are developing at an unprecedented pace. Data collection, analysis, interpretation, presentation, review, and publication take place completely on computers. The final product often is still a static document with only limited links to the underlying digital material, making transparency and reproducibility a challenge. In this work we apply the mechanism of badges to provide prominent connections to underlying analyses environments and important (meta-)data to readers of scholarly publications in geospatial data science. An API specification and implementation for a badge server provide extended and regular badges. The badges leverage recognition value for executability, licensing, spatial extent, and peer-review metadata – base information which either is or should be made available. We show a client-side integration method across many third-party platforms that allows evaluation of badges in realistic scenarios. The server and client enable an independent spreading of badges. The open source publication of all used software enables reproducibility and extensibility. The badges show potential to enhance interaction with scholarly works and should be evaluated with academic users in the future.

Open is not enough

“…Besides promoting and motivating scientific practices, science badges are expected to improve the discoverability of building blocks of research [139]. Through tailored promotion of research content and uses, we aim to support navigation and discovery of preserved research.…”

“…The targeted exposure of scientific building blocks is expected to provide an effective means to increase the visibility of scientists and research work, which in turn motivates contributions. Given the impact on the visibility of research and scientists' careers, we acknowledge that the provenance of science badges plays a crucial role in establishing researchers' trust [139].…”

The role of metadata in reproducible computational research

“…For example, the Journal of Water Resources Planning and Management waives or reduces fees of reproducible papers, 227 and several journals promote reproducible works, e.g., through badges. 228 Other carrots could involve a support mechanism by which pre- and post-publication peer review can properly evaluate and test statistical methods cited in papers. Such a collaborative computational peer review could involve parameter exploration, swapping out individual statistical tests or tool components for similar substitutes, and using new datasets.…”

Serverless GEO Labels for the Semantic Sensor Web

“…With the practical solutions for label generation introduced in this work, the actual spreading of labels will require leading organisations to add and maintain labels on their widely used geospatial catalogues. In the meantime, a bottom-up approach with client-side label integration [23] could provide the benefits of GEO labels to interested users, and the GEO label can be examined in relation to recent developments on scientific data publication such as the FAIR Guiding Principles [26].…”