Progress update 2 – metadata mapping
Mapping of current UK Data Archive metadata to geospatial standards: DDIlight and INSPIREation (sorry)
The UK Data Archive currently follows a DDI2 metadata schema, with a view to move to DDI3 in the near future. Such changes inevitably involve a certain amount of upheaval, and one of the main goals of this project is to ensure that in the process the Archive comes in line with current geospatial standards. What’s the point you ask? Well, the result is metadata which is more widely interpretable by both the human and machine reader – be that a researcher trying to get his/her head around a particular variable, or a search algorithm scouring a catalogue for a particular value.
The aforementioned ‘current geospatial standard’ is the EU directive INSPIRE. GEMINI is the UK adaptation of this – specifically tailored to UK practitioners. I have spent a chunk of my time recently putting together a metadata ‘mapping’ i.e. a comparison of the various schema involved. The result will highlight additions needed and contribute to the Archive metadata roadmap.
Some of the additions required are fairly straightforward, such as better documentation of dataset spatial resolution. Others though are more tricky – how, for example, do we allow for GEMET keywords (GEMINI’s thesaurus of choice) in a closed, single vocabulary system?
Perhaps the greatest challenge is deciding exactly how far to go i.e. can you future-proof to any extent? A good example of this is in the bounding box storage. A bounding box is useful as a very simple method of recording a geographic area – only four variables need to be stored: east, west, north and south. Isn’t this a bit limited though? Wouldn’t a bounding polygon be much better? What if a depositor submits this information but we have no framework to store it? The chances are that this kind of information as part of a data deposit will become more common. What this all comes down to is that while implementing metadata is complex and the specifications
painfully restrictive necessarily exacting, we can’t just set out to tick a series of boxes.