 |
 |
Hmm, just because a given page is linked to from a wikipedia page, can one assume that the target of the link is about the same thing as the original page? I'm not sure how often this assumption would be violated? Ed Summers wrote: > On Mon, Mar 29, 2010 at 3:45 PM, Lars Aronsson <[log in to unmask]> wrote: > >> Wikipedia has 720 links to www.rockhall.com, >> which is a dream for any website, >> http://en.wikipedia.org/wiki/Special:LinkSearch/www.rockhall.com >> > > I had no idea that Wikipedia provided a live search of external links. > Up until now I have been pulling those relations from our referrer in > our web server logs. > > If you squint right (with your Linked Data tinted glasses on) > wikipedians sure look like catalogers, describing the resources that > libraries and museums put on the web. > > For example from > http://en.wikipedia.org/w/index.php?title=Special:LinkSearch&target=http://chroniclingamerica.loc.gov&limit=500&offset=0 > it's possible to infer: > > <http://chroniclingamerica.loc.gov/lccn/sn83045211/1915-02-08/ed-1/seq-16/> > dcterms:subject <http://dbpedia.org/resource/Constance_Drexel> . > > Since we don't have page level metadata for the 2 millions pages of > newspaper content in Chronicling America, this is pretty valuable > information. > > //Ed > >