Feature Request: Support supplementary dataset metadata and optional dataset assessment

[cmbz]

Overview

• Data repositories have historically emphasized support for general and domain-specific bibliographic, descriptive dataset metadata (e.g., author, publisher, geolocation, sample size, etc.). This metadata tends to be assigned by data depositors, data owners, and/or data curators.
• However, increasingly, data producers and human and machine data users have expressed the need for associating datasets with supplementary, ad hoc metadata to support reuse, computational processing, and searching for data that meets specific criteria. This metadata is distinct from Dataverse's standard or custom metadata blocks.

This functionality will enable installation and repository managers to:

1. Define and set supplementary metadata elements and value ranges (setting via UI is not required)
2. Define the scope of the metadata (e.g., installation, collection)
3. Supply optional plug-in dataset assessment functions
4. Activate the assessment function so it runs periodically on all datasets in scope (collection or installation)
5. Make supplementary metadata (and assessment scores) available in search
6. Optionally display the metadata in the Dataverse UI
7. Optionally download/export or share the metadata

Use Cases

• Harvard Dataverse Acceptable Dataset Characteristics
• FAIRSCAPE AI Readiness metadata and scoring
  
• MIDAS 2.0: Assessing health dataset readiness for safe, trustworthy AI
• Croissant Responsible AI
• Administrative metadata (e.g., metadata that is repository-specific, workflow-specific, reflects general records schedule status)
• AI-enhanced metadata (e.g., GREI: AI Supplement #453 and Project: Enhancing research data metadata using generative AI #437)
• Related Datasets (also: Connecting research outputs in Dataverse)

Other Examples

• Kaggle Usability Scores

Deliverables

• Implementation-ready technical design document that has had input from the Dataverse Community and Dataverse Core team
• Note that the design document may also recommend approaches for specific use case categories.

Resources

• Google Drive Folder