Building Linked Open Greek Pottery

I have been working since the fall on a side project to apply the methodologies learned in the development of, a collaborative project that seeks to define all of the intellectual concepts of numismatics following linked open data methodologies, toward the development of a similar resource for ceramics. was registered a few months ago, but it is still very much a prototype. I plan to do a demo of the project in two weeks at CAA in Paris.

When I began the project back in October or November, I started by scraping the HTML thesauri at the Beazley Archive at Oxford and comparing terms to the British Museum thesauri in PHP. This was really complicated, and I’m not sure I was using the most efficient tools for this job. Instead of getting bogged down in this process, I decided to switch gears (and do what I normally do), and create a few dozen really well-defined RDF identifiers for Greek pottery concepts and build an entire scalable framework on creating, editing, and publishing identifiers (complete with the XForms backend and HTML5 public user interface), as well as develop workflows for ingesting and managing RDF data dumps of physical vases in museum collections.

Not long ago, I received a CSV dump of 24 vases from the Getty Museum, which I just now mapped into a hybrid CIDOC CRM/ ontology (though leaning heavily on existing CRM classes and properties). I posted these data into the SPARQL endpoint a few moments ago. You can take a look at basic query results:

Derived from the query:

PREFIX rdf:    <>
PREFIX dcterms:    <>
PREFIX skos:    <>
PREFIX owl:    <>
PREFIX foaf:    <>
PREFIX ecrm:    <>
PREFIX geo:    <>
PREFIX kid:    <>
PREFIX kon:    <>

SELECT ?s ?title ?identifier ?technique WHERE {
?s a ecrm:E22_Man-Made_Object .
?s dcterms:title ?title .
?s dcterms:identifier ?identifier .
?s ecrm:P32_used_general_technique ?turi .
?kturi skos:exactMatch ?turi .
?kturi skos:prefLabel ?technique
FILTER ( lang(?technique) = "en" )
} LIMIT 100

By this time next week, I should have ingested several dozen vases from the British Museum into the endpoint so that I can start building visualization mechanisms that take advantage of data from both museums. For example, there will be SPARQL-based maps and quantitative analyses rendered in charts and graphs (the sort of thing that has been implemented in OCRE and Coin Hoards of the Roman Empire).