Sanda:
There are a few examples, but one big question is what you want to use it for. DDI-CDI can do a lot of different things, the two major cases being process description and data integration.
One project which is very well documented uses DDI-CDI to describe the process of integrating data (the data itself is described using Lifecycle). This is the "Climate Neutral and Smart Cities" project, which combines environmental and air quality data with the European Social Survey. You can find all of it at
https://www.europeansocialsurvey.org/esslabs .
I think Sikt has done more here in a continuation project called SoGreen - Hilde or Benjamin from Sikt could fill you in.
Another use case is the Cholera and Flooding Disaster Risk Reduction project in Malawi. Here, data from the public health authority about cholera deaths and vaccinations is combined with climate and flooding data to help with disaster preparedness and response. (The flooding leads to problems with the water supply, which causes cholera outbreaks, basically - what is termed a "cascading disaster".). Here, DDI Codebook and SDMX are being used as inputs to produce DDI-CDI, which allows for integration of the data, and is used for dissemination as part of a CDIF-compliant FAIR package held in Dataverse. Here, this becomes the basis for producing Croissant ML for consumption by AI agents.
This project is still being implemented, but Chifundo Kanjala gave an excellent presentation about it in late May to the UNECE Stats Modernization meeting, so maybe you could get ahold of his slides. (Let me know if you need help with this - I am on the project team for CODATA doing DDI-CDI and SDMX support).
Finally, it might be worth reaching out to Darren Bell or Deirdre Lungley at UKDA - they have implemented DDI-CDI for their variable shopping-cart application - I don't know the production status.
Another use case is the CDIF4XAS project, where they are prototyping DDI-CDI as an exchange format for X-Ray Absorption Spectroscopy data. This is an interesting case, because they have two different domain standards, one using a text-based data format, and the other one in HDF5. They are using DDI-CDI as the way of describing the data sets for exchange between applications and institutions.
There are a bunch of other implementations, but those are the ones I am most familiar with right now.
Like a lot of RDF implementations, we mix and match metadata fields from lots of different standards.
We just launched a huge 3-year project called CDIF4EOSC which will do use cases in climate adaptation, social sciences, and material sciences. COADATA is leading the project, and Sikt and UKDA are involved (along with lots of others). This project will further develop CDIF as the set of standards for use in European Data Spaces and EOSC Nodes, which means there will be a lot of DDI-CDI in use within EOSC and Europe generally.
Anyway - I hope that helps!
Cheers,
Arofan