Now you can check out the project description for our Civil War Pilot project #1 to give you some idea of what we’ve been up to.
We’re sort of starting in the middle with this blog (well, close to the end, actually, at least for the first “pilot”), but I hope to give you some background of our experience and the flavor of the project.
This one is an interesting stew… a first try at transforming, creating, modeling, storing, querying and displaying linked data using open source tools. We had a lot of ambitions as we were planning the pilot. Some of us suspected we could only bite off small pieces of these things in 3 months, and our experience is bearing this out! But we’re learning loads.
Since the proposal, we’ve been fortunate to add library Fellow Bethany Nash to the group. Today she and I sat down and worked through creating some “original” triples describing some resources we don’t really have metadata for.
We are starting simple (using Dublin Core terms) with the idea that we will branch out to other vocabularies as we see the need, and have already come across some interesting questions (for a letter, should a dcterms location field be created for both the place the letter was written, and a different place that the letter talks about?)
After having no luck getting someone with time to build a web input form (yet), and after trying to use Excel (not a good move; too hard to get it to stop reformatting things like URLs!), we’re just creating N3 triples in a MS Word Document, validating them with the rdf:about Validator and Converter, and then pasting them into the Sesame OpenRDFWorkbench “Add” input box. N3 is way easier to work with than RDFXML. E.g.:
@prefix dcterms: <http://purl.org/dc/terms/>.
<https://duchamp.library.emory.edu/omeka/admin/items/show/123> dc:title “Letters, June 17, 1864, ‘near Atlanta'”.
Leave a Reply