BrightSpots Get Brighter
/Ted Habermann, Metadata Game Changers
Introduction
During June we discussed results of FAIR metadata assessments for 387 DataCite repositories associated with universities and colleges. The University of Bath Research Data Archive was identified as an outstanding example of complete DataCite metadata supporting four FAIR use cases (findable text, findable identifiers, connections, and contacts).
After that blog was published, a research librarian at the University of Bath, Alexander Ball, let me know that they were in the process of updating their tools for managing people and organizations in their metadata, both important elements in the FAIR use cases. Last week he let me know that they had finished that project and updated their DataCite metadata. Thanks Alex for a great opportunity to see how improvement iterations are reflected in the FAIR Assessments.
January Baseline
Figure 1 shows the results of the initial assessment done during January 2024 as completeness of 58 metadata elements in four FAIR use cases (Findable Text, Findable Identifiers, AIR Connections, and AIR Contacts). The scores for each use case are shown above the plots on the left and in a line plot on the right along with the total score of 46%. This was the highest total score observed in the 387 repositories studied.
Figure 1. University of Bath FAIR Assessment - January 2024. Radar plots on the left show completeness for 58 metadata elements in four FAIR use cases. Use case scores are also shown on the right. The total score of 46% was the highest observed in 387 repositories.
September
Figure 2 shows the same data collected during September 2024, after the University of Bath completed improvements to their people and organization identifiers and updated their metadata. Improvements of over 30% are apparent as large increases in the Findable Identifiers and AIR Contacts use cases (increases in the blue area in the radar plots and shown by arrows in the right frame).
Figure 2. University of Bath FAIR Assessment – September 2024. Radar plots on the left show completeness for 58 metadata elements in four FAIR use cases. Use case scores are also shown on the right. The total score increased to 59% while the use case scores for findable text and AIR contacts (green) each increased by more than 30%.
Figure 3 shows the details of the completeness improvements accomplished by this repository in the Findable Identifiers and AIR Contacts use cases (light green in Figure 2). The focus on identifiers for people (ORCIDs) and organizations (RORs) are clear along with the inclusion of the identifiers for the DataCite contributorType RightsHolder. Smaller increases also occurred in the free-text resourceType and the resourceIdentifierType elements.
Figure 3. Details of metadata improvements at the University of Bath.
Conclusion
Identifiers for individuals and organizations are critical metadata elements for finding and connecting research objects of many kinds. The University of Bath Research Data Archive metadata now include ORCIDs identifying 738 researchers occurring 1883 times and 180 RORs occurring 2355 times. The RORs connect the University of Bath and 179 collaborating organizations to resources held in the archive. The repository metadata also includes other important recommended and optional elements, e.g. contributors of many types (Researcher, Supervisor, DataCollector, and others), funder and award identifiers, and methods descriptions. Congratulations to the University of Bath Library Research Data Archive for providing such great examples!
Note: Finding RORs for organizations can be a daunting task without tools. If you need help finding RORs for organizations in your metadata, take a look at RORRetriever.