documentation for entity import process and supported data / formats

[rlskoeser]

We're trying to install and evaluate EATS for possible use with a project, and the import/export documentation isn't very clear. What import formats does EATS support? Is it only the custom EATSML that the exporter generates; is there any documentation for that format (i.e., could it be generated from a non-EATS system)? In particular, I'm interested to know if it would be possible to import data from VIAF and use VIAF entities and ids in an EATS system (or does that sound like the wrong approach / a bad fit for EATS?).

[ajenhl]

Currently the only import format supported is EATSML. EATSML can certainly be generated from a non-EATS system. From a quick look at a VIAF XML record, the mapping shouldn't be too difficult. You can either have VIAF itself as your sole authority and put the name or names you want, and a subject identifier assertion pointing to the VIAF record, or you can make each each participating institution an EATS authority, along with VIAF.

As for whether this is a good approach, I'm not sure - it depends on what you want to use EATS for.

There isn't really any documentation on the format itself, beyond the RelaxNG schema (in server/lib/eatsml.rng). It adheres closely to the EATS model, so entities are made up of property assertion elements, specifying who says what about the entity. I'd be happy to write more documentation if I knew what information would be useful.

[rlskoeser]

Thanks for pointing me to the EATSML schema, I hadn't discovered that yet. Is the EATS model you describe documented anywhere? I've glanced at the papers linked from the googlecode site; it would be helpful if you added an overview of the data model to your sphinx docs and/or doc strings to your db model objects.

We've been trying to get a test instance of EATS set up to get an idea of what it does and if it might be useful for us. If we import VIAF into EATS, would it be possible to use the Oxygen plugin to update XML based on VIAF ids, or does the system only work with internal EATS ids ?

[ajenhl]

The best conceptual overview of the EATS model is in the Introduction to EATS paper. It deals with the old EATS, and some details have changed, but the basic idea of property assertions and authorities remains the same.

The system does only work with internal EATS IDs, which is not ideal. It would be nice to allow for any of an entity's subject identifiers to be used, with the client tools querying the server for the entity based on that. Of course, this does raise its own issues, particularly the need for an extra step and UI for selecting which subject identifier to use, where there is more than one, when selecting an entity in the client.

Since EATS provides a store of mappings between its IDs (which are URLs) and other URLs, it's not an enormous burden to have the EATS URLs stored in XML documents. If these documents need to be processed by other systems that expect a VIAF URL, then they can be transformed (in both directions) easily enough on the basis of the data in an EATSML export. A little tedious, but I think that's guaranteed regardless (though I'd love to be shown a better way).