W3C home > Mailing lists > Public > public-vocabs@w3.org > October 2011

Schema.org-related experiment with Wikipedia/authority files

From: Mathias Schindler <mathias.schindler@wikimedia.de>
Date: Mon, 24 Oct 2011 02:14:03 +0200
Message-ID: <CA+LWsp3eEc0huH=JKkCw4o=xcg5fYSmTi=zSNY37ocMzmoZG_g@mail.gmail.com>
To: public-vocabs@w3.org
Greetings,

I would like to draw your attention to an experiment we are currently
conducting over at http://toolserver.org/~apper/dpa/index.php (it is
in German only).

The German news agency dpa kindly provided us with 32.000 wires from
May 2010 for experimentation purposes. We are evaluating
semi-automatic tools to tag names in these wires and to link the
specific persons to the authority file of the German National Library
and to the list of persons of the German language Wikipedia. A version
of these tools has already been used in a collaboration between
Wikimedia Germany and the German Federal Archives
(http://toolserver.org/~apper/ba/).

Once we have established a link between a name in a dpa wire and a
specific person at one of those two sources, we can automatically pull
various data out of it and embed this information in schema.org
compliant manner. A very early and not yet functional demo can be seen
at http://toolserver.org/~magnus/granDPA/index.php?id=6429 (you might
want to have a look at the source)

Third party external resources using the German authority file PND
(Personennamendatei) are linked to via the PND Beacon system
(http://de.wikipedia.org/wiki/Wikipedia:BEACON - again, German, an
English summary exists at http://meta.wikimedia.org/wiki/BEACON)

One aim among many of this exercise is to evaluate the coverage of
German language biographical articles in Wikipedia over persons
mentioned in news agency coverage.

I appreciate any comment you may have.

Mathias

-- 
Mathias Schindler
Projektmanager
Wikimedia Deutschland e.V.
web: http://www.wikimedia.de
Received on Monday, 24 October 2011 21:10:38 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 22 May 2012 06:48:56 GMT