Investigation of the Persistence of Terminologies’ Terms

Terminology terms can be used to annotate data in a machine-readable and understandable way. Keywords describing datasets or publications can be annotated with the URLs of the corresponding terms in a vocabulary or ontology. Unfortunately, this annotation breaks if the corresponding URLs are no longer resolvable. This might happen if a term is deleted - e.g. in a new version of a terminology, where this term has not been only marked as deprecated, but has been deleted. We therefore investigated whether this deletion and/or replacement of terms had happened before in the terminologies of the Earth System Science (ESS) collection, provided by the TIB Terminology Service.

Specifically, we analysed the persistence of classes and individuals of 32 terminologies in the ESS collection that we could compare with previous versions. We aimed for a qualitative rather than a quantitative analysis, i.e. we only examined successive earlier versions of a given terminology until we found deleted classes or individuals. Where appropriate, we either compared different versions of a term or analysed the timeline, tracking changes. Additional checks revealed that the corresponding URLs were no longer resolvable. More information on this study can be found in a report by A. Ganske (2024).

Some changes were found in very early versions of the respective terminology, which may not have been used to annotate the data. Also, FAIR terminology issues and practices have evolved over the years; hence, as communities adopt best practices there is a high probability that fewer classes will be deleted in future versions. However, the risk that terminology terms won't be available after some time is likely to remain.

As a result of this investigation, we decided to carry out a feasibility study: can we assign persistent handles to all classes and individuals of each future version of a terminology in the TIB TS ESS collection, so that the handles redirect to the landing page of the respective terms in the TIB TS? These handles could then be used in annotations, and the TIB will ensure that they are persistent. This is work in progress - we will report as soon as we have new results.