W3C home > Mailing lists > Public > public-vocabs@w3.org > June 2012

schema.org/sameThingAs proposal

From: Dan Brickley <danbri@danbri.org>
Date: Fri, 22 Jun 2012 17:21:24 +0200
Message-ID: <CAFfrAFqoZhb1sSJTBUNqLwkuR0y4+OCJvz1UrK4v00jYKxRE3Q@mail.gmail.com>
To: public-vocabs@w3.org
I've written up a proposal at
http://www.w3.org/wiki/WebSchemas/ThingIdentity  for a new schema.org
property, sameThingAs. In fact it's been there a while; I thought this
mail was sent already!

"This is a proposal to improve and clarify schema.org's handling of
identity issues, in particular for the common case where diverse sites
provide information about the same real world entity. It adds a
property to schema.org, 'sameThingAs' that can be used to indicate
when a single real-world entity is being described." (see wiki for
lots more detail)

The idea is similar to W3C's existing 'owl:sameAs' relationship, but
more flexible / loose, since with schema.org markup (and
rdfa/microdata in general) it is very common to see mixing of
identifiers for things, and identifiers for pages-about-those-things.

Basically all sameThingAs says is, "whether these are direct or
indirect identifiers doesn't matter for now; they point to the single,
same real-world entity".

This topic has a lot of historical baggage in the standards world;
I've put a bit of that in the wiki post. The W3C TAG (tech
architecture group) are also reconsidering their view of related
topics, specifically the question of using #-less, non-redirecting
http://* URIs to identify 'real world' entities.

Traditionally this sort of topic tends to result in gigantic and
unproductive email threads, so I post this with some caution, and
encourage contributors to consider commenting in the Wiki instead of
email. There is a little bit of admin to do to get a W3C account for
Wiki use, but they're available to everyone from
http://www.w3.org/Help/Account/Request/Public ... Mailing list
comments are easily lost, whereas notes in a Wiki page can be
integrated, improved and synthesised.

For a motivating scenario, see IMDB's current schema.org descriptions
(copied here from the Wiki page),

<h4>Stars:</h4>
<a href="/name/nm0010930/"   itemprop="actor">Douglas Adams</a>,
<a  href="/name/nm0048982/"   itemprop="actor">Tom Baker</a>
and <a  href="/name/nm3035100/"   itemprop="actor">Hans Peter Brondmo</a>
</div>

...this from a page about a film, with links to pages about
people/actors, where the person was an actor in the film. Our
motivating scenario here is: how can we improve the merging of data
about the actors without making a giant mess of otherwise simple
markup? In other words, find ways to use the 'person page' as a
stand-in identifier for the person. And (tricky part...) what if
anything can we do when indirect identifiers such as Wiki pages are
used to identify 'real world' things that are also 'creative works'?
e.g. http://en.wikipedia.org/wiki/Steamboat_Willie

Thanks for your thoughts,

cheers,

Dan
Received on Friday, 22 June 2012 15:21:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 22 June 2012 15:21:56 GMT