Testing Schema.org output formats

Part of my efforts working with Google in support of the Schema.org structured web data vocabulary, its extensions, usage and implementation, is to introduce new functionality and facilities on to the Schema.org site.

I have recently concluded a piece of work to improve accessibility to the underlying definition of vocabulary terms in various data formats, which is now available for testing and comment. It is available in a prerelease of the next Schema.org version hosted at webschemas.org – the Schema.org test site.  It is the kind of functionality could benefit from as many eyes as possible to check and comment.

Introduced in the current release (3.1) was the ability to download vocabulary definition files for the core vocabulary and its extensions in triples, quads, json-ld and turtle formats.  Building upon that, the new version, provisionally numbered 3.2, introduces download files in RDF/XML, plus CSV format for vocabulary Types, Properties and Enumeration values.

There have been requests to make other output formats available for individual terms in addition to the embedded RDFa already available.  In response to these requests, we have introduced per term outputs in csv, triples, json-ld, rdf/xml and turtle format.

These can be accessed in two ways. Firstly by adding the appropriate suffix (.csv, .nt, .jsonld, .rdf, .ttl) to the term URI:

• http://webschemas.org/Book.ttl
• http://webschemas.org/Product.rdf
• http://webschemas.org/description.jsonld
• http://webschemas.org/Paperback.nt
• http://webschemas.org/LocalBusiness.csv

They are also available using content-negotiation, supplying the appropriate value for Accept in the httpd request header (text/csv, text/plain, application/ld+json, application/rdf+xml, application/x-turtle).

So if you are interested in such things, take a look and let us know what you think about the contents, formats, and accuracy of these outputs.