[Web4lib] Authority files and Wikipedia

Lars Aronsson lars at aronsson.se
Sat Aug 13 17:02:00 EDT 2005


Last weekend the Wikimania conference was held in Frankfurt, 
Germany, not at the Sheraton or Marriott but at a youth hostel.  
This was the first international conference devoted entirely to 
the development of Wikipedia (the free encyclopedia) and related 
projects.

One of the presentations was given by Christina Hengel and Barbara 
Pfeifer about "VIAF: Linking the Library of Congress and OCLC 
records with Personendaten".  As some web4lib subscribers probably 
know already, VIAF is short for Virtual International Authority 
File and is an OCLC-assisted collaboration between Die Deutsche 
Bibliothek (ddb.de) and the Library of Congress aiming to 
synchronize the name authority files of the two national 
libraries.

This presentation was given in light of the existing cooperation 
between DDB and the German language branch of Wikipedia, hinting 
at how this could be expanded to other languages and libraries.  
This cooperation was also featured in a press release from DDB on 
August 2, "Wikipedia nutzt Katalog Der Deutschen Bibliothek und 
deutsche Normdateien".  I fail to find a bookmarkable URL for this 
press release, but you can find it under "News" at www.ddb.de.

The way the German cooperation works is that every name authority 
has been given a unique ID number by DDB, and this number can be 
used in a URL to link to the DDB catalog.  The German name 
authority file (PND = Personennamendatei) contains 2.6 million 
names for 600,000 people. For example the German writer Thomas 
Mann (1875-1955) has been given number 118577166 and this is 
encoded as {{PND|118577166}} in the editable text of the 
biographical article http://de.wikipedia.org/wiki/Thomas_Mann
creating a link when the page is saved to 
http://dispatch.opac.ddb.de/DB=4.1/REL?PPN=118577166

There is nothing magic in this of course, except that the 
volunteers of the German Wikipedia have gone through all 270,000 
articles and found 38,508 biographies, and matched no fewer than 
14,013 of them to the DDB authority file and inserted the PND 
numbers in these Wikipedia articles.

This Wikipedia effort was described in another presentation by 
Jakob Voss, "Metadata with Personendaten and beyond",
http://en.wikibooks.org/w/index.php?title=Wikimania05/JV2

It is now trivial to compile a list of PND numbers and the 
matching Wikipedia URLs, which could be useful for the DDB catalog 
if it wanted to link back, which we have yet to see.  Commercial 
websites such as Amazon.de could just as easily look up an ISBN 
number in DDB, find the PND number of the author, and link to the 
author biography in Wikipedia.

The German branch is the 2nd biggest in Wikipedia, surpassed only 
by the 680,000 articles of the English language Wikipedia.


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se


More information about the Web4lib mailing list