Ofgem published its decision on updates to it’s Data Best Practices Guidance v1 on August, 7 2023. Regulated companies in the UK energy sector are required to follow these guidelines as part of their RIIO-2 license obligations. It also sets a bar for data sharing in the wider energy sector, and will almost certainly influence guidance in other regulated sectors such as water.

A particularly interesting update for those of us aiming to improve interoperability and accelerate data sharing was the specification of Dublin Core as the minimum metadata standard, with a requirement for it to be implemented within 12 months i.e. August 6, 2024.

Decision on updates to Data Best Practice Guidance and Digitalisation Strategy and Action Plan Guidance

3.35. 71% of respondents approved of the proposal to mandate Dublin Core as a common Metadata standard, with an additional 21% approving but with queries or caveats. When asked for alternative suggestions for a common standard, the only option considered was DCAT, a variant of Dublin Core.
3.36. We have considered this, weighing its merits against the established and tested nature of Dublin Core, and the international and multi-sector uptake of Dublin Core as a standard, and have decided to mandate Dublin Core as the Metadata standard.

In our response to the consultation that led to Ofgem’s update we advocated for the standard to be DCAT, a refined and extended version of Dublin Core that is widely supported by data catalogue software, and in the form of DCAT-AP is the primary format in the EU for government data catalogues and the organisations that report into them.

DCAT and CKAN are mostly 1:1 compatible with the 15 Common Fields in the Dublin Core but there are some gaps. DCAT doesn’t include the coverage, date and subject fields, and the standard CKAN DCAT plugin doesn’t import or export the creator, contributor and relation fields.

Icebreaker One has searched for definitive guidance for mapping these gaps between DC and DCAT (these mappings are called “crosswalks”). We didn’t find anything that answered the question once-and-for-all, but there are some clues.

  • The Research Data Alliance built a crosswalk into schema.org from various standards, including DC and DCAT. The data behind this is in a CSV on the RDA GitHub. By seeing how they think DC and DCAT should map to schema.org, we can infer what a mapping between the two would mean.
  • The EU is putting a lot of effort into this at the moment under the European Open Science Cloud (EOSC) initiative. Specifically a report on interoperability that proposes a lingua-franca for metadata in Appendix I, and the work to establish a crosswalk registry between standards, which is still in the requirements-gathering stage. We’re looking forward to this work reaching determinations that will help drive universal crosswalk standards.

In the absence of clear external standards, we have developed a proposed mapping – see below.

We would welcome feedback on this. You can leave comments directly in the document.