Day three of the Semantic Tech & Business Conference in San Francisco brought us a panel to discuss Schema.org, populated by an impressive array of names and organisations:
Ivan Herman, World Wide Web Consortium
Alexander Shubin, Yandex
Dan Brickley, Schema.org at Google
Evan Sandhaus, New York Times Company
Jeffrey W. Preston, Disney Interactive Media Group
Peter Mika, Yahoo!
R.V. Guha, Google
Steve Macbeth, Microsoft
This well attended panel started with a bit of a crisis – the stage in the room was not large enough to seat all of the participants causing a quick call out for bar seats and much microphone passing. Somewhat reflective of the crisis of concern about the announcement of Schema.org, immediately prior to last year’s event which precipitated the hurried arrangement of a birds of a feather session to settle fears and disquiet in the semantic community.
Asking a fellow audience member what they thought of this session, they replied that the wasn’t much new said. In my opinion I think that is a symptom of good things happening around the initiative. He was right in saying that there was nothing substantive said, but there were some interesting pieces that came out of what the participants had to say. Guha indicated that Google were already seeing that 7-10% of pages crawled already contained Schema.org mark-up, surprising growth in such a short time. Steve Macbeth confirmed that Microsoft were also seeing around 7%.
Another unexpected but interesting insight from Microsoft was that they are looking to use Schema.org mark-up as a way to pass data between applications in Windows 8. All the search engine folks were playing it close when asked what they were actually using the structured data they were capturing from Schema.org mark-up – lots of talk about projects around better search algorithms and indexing. Guha, indicated that the Schema.org data was not siloed inside Google. As with any other data it was used across the organisation, including within the Google Knowledge Graph functionality.
Jeffrey Preston responded to a question about the tangible benefits of applying Schema.org mark-up by describing how kids searching for games on the Disney site were being directed more accurately to the game as against pages that referenced it. Evan Sandhaus described how it enabled a far easier integration with a vendor who could access their article data without having to work with a specific API. Guha spoke about a Veterans job search site was created with the Department of Defence as they could constrain their search only to sites which only included Schema.org mark-up and identified jobs as appropriate for Veterans.
In questions from the floor, the panel explained the best way of introducing schema extensions, using the IPTC rNews as an example – get industry consensus to provide a well formed proposal and then be prepared to be flexible. All done via the W3C hosted Public Vocabs List.
All good progress in only a year!
Richard Wallis is Technology Evangelist at OCLC and Founder of Data Liberate