Usage Based Discovery (UBD) is a new paradigm for dataset discovery that depends on the broad implementation of open science best practices across a variety of Earth Sciences domains. Scientific journals and other research outlets promote the reproducibility of research results by exposing the data that were used to create scientific analyses and data products. Over the past three ESIP federation meetings, our cluster has held hackathons focused on harvesting dataset usage relations from these sources. The Cluster has also collaborated on a graph based data discovery and visualization framework to expose the information that we have collected. These activities allowed us to explore and understand the UBD concept, and we think this new paradigm promises to improve the data management enterprise and reach a broader user base, thereby maximizing the benefit of the Earth Sciences to society.
- Introduction - 15 min
- Introductions and session overview - 5 min
- Usage Based Discovery Cluster Accomplishments and Next Steps - 10 min
- Invited Talks - 30 min
- Publink (Madison Langseth) - 15 min
- Coleridge Initiative (Julia Lane and Gizem Korkmaz) - 15 min
- Sense Making [Slido Poll] includes selection of breakout topic so we can decide how many breakout rooms to open. - 5 min
- Future of UBD at ESIP Breakout Discussion - 35 min:
Participants will self organize into breakout session and discuss their breakout topic of choice for the next 25 minutes, following the What? / So What? / Now What? format (see liberating structures). Then with the remaining ten minutes, we’ll regroup and share report outs of what the breakouts discussed. - 25 min- Messaging Usage Based Discovery utility and benefits.
- User interface functionality that show users the benefits of UBD discovery.
- How to win seed funding and enroll participants in cluster.
- Share ideas, best practices, and help develop community standards.
- DOIs - structuring Digital Object Identifiers to facilitate dataset discovery?
- Sharing Graph Data - How to share UBD data generated by different projects?
- Crowdsourcing UBD - how to get users to report mistakes, make suggestions for other pubs, and earn trusted roles as expert UBD data curators?
- UBD analytics, content, ML, and UI:
- Topics and findings - ways to use the data citation graph? Other research questions?
- Data Development - Enrolling people to add their dataset-usage relations to the UBD Graph DB/ UI.
- Machine learning - how to detect informal data citation.
- UBD Data Architecture - technology behind the UBD graph database and allied technologies such as the semantic web and google/schema.org.
- Regroup and report out: breakout groups select one participant to share what they discussed.- 10 min
- Adjourn the meeting - 5 min
View Recording
View Notes
