Hi
I am interested in using calais to generate dbpedia/linked data concept extractions from news articles. I have been experimenting with the REST api, to perform a calais concept extraction, and then re-requesting json formatted responses for the main calais terms extracted, to pick up the same-as linked data.
However, not many (if any) of the calais terms produce same-as data, even though dbpedia entries doe xist in dbpedia.
For example, a concept extraction on a BBC sports artical about Andy Murray, generates a number of calais terms , ranked accordingly (this is all teally good!), eg:
Andy Murray: http://d.opencalais.com/pershash-1/ae42bfe1-508c-3b0e-9c1c-d22ba48643e9
Roger Federer: http://d.opencalais.com/pershash-1/6b9485f5-dc32-3879-b0d4-cce73c363a0a
Now, both Andy Murray and Roger Federer both have dbpedia topics, but when the Calais terms are requested (as JSON):
http://d.opencalais.com/pershash-1/ae42bfe1-508c-3b0e-9c1c-d22ba48643e9....
http://d.opencalais.com/pershash-1/6b9485f5-dc32-3879-b0d4-cce73c363a0a....
neither produce same-as links to their dbpedia equivalents ?
So, am I missing something, or is the linked-data querying very sparse and imcomplete with respect to dbpedia ?
Thanks for any help
Paul

Are these issues been resolved?
Web development and Website Development
Hi all,
I have the same use case as Paul (who started this thread). Like him, I also get a JSON with the annotations, grab all entities of interest, then dereference them in order to obtain DBpedia sameAs links. But it seems I'm going the longer route here.
For example, the entity URIs I get from the annotation service are like this:
IBM
http://d.opencalais.com/comphash-1/7c375e93-de13-3f56-a42d-add43142d9d1
When I try to dereference that, I get a proprietary non-standard (non-HTTP) redirect:
<cld:redirection rdf:resource="http://d.opencalais.com/er/company/ralg-tr1r/9e3f6c34-aa6b-3a3b-b221-a07aa7933633"/>
Connecting to that URI I can see a couple of sameAs (including one to paid content), but none to DBpedia
<owl:sameAs rdf:resource="http://cb.semsol.org/company/ibm#self"/>
<owl:sameAs rdf:resource="http://p.opencalais.com/er/company/ralg-tr1r/9e3f6c34-aa6b-3a3b-b221-a07aa7933633"/>
What's happening? Maybe the bug I observed in the interface is actually in the server-side?
Why so many indirections?
Any chances you guys will make this process a bit friendlier?
Cheers,
Pablo
Hi all,
I have also noticed that sometimes the web page for an entity "hides" the sameAs link to DBpedia, or the Aliases. Reloading the page sometimes gives different results. Try reloading the page many times and you'll notice the behavior.
Looks like a bug in the interface?
Pablo
Hi Pablo,
This definitely sounds like a bug. We will take a look at the logs, meanwhile can you provide us the link where you see this behavior?
thanks
sumit
The one for IBM, from the documentation:
http://d.opencalais.com/er/company/ralg-tr1r/9e3f6c34-aa6b-3a3b-b221-a07aa7933633
Got the link from here.
http://www.opencalais.com/documentation/linked-data-entities
Using Mozilla Firefox 3.6.11 on Windows Vista.
The one for Rio Tinto is working randomly (need to refresh the page several times to get the sameAs :
http://d.opencalais.com/er/company/ralg-tr1r/ec3b7ab9-15f5-31a3-8231-6cba75adeb1a.html
Using Mozilla Firefox 3.6.12 on Ubuntu.
Any news on this bug?
Looks quite annoying :)
Hi all,
I am using this thread to ask a question you should obviously have the the answer. In your examples given in http://www.opencalais.com/documentation/linked-data-entities the links to the LinkedData (sameAs to freebase or dbpedia) just disapeared. Even for the examples (IBM) these sameAs links are not there anymore.
Is it a bug? Is there a change in the API I did not see?
Thanks in advance!
Hi Paul,
We are expanding the list of entities we have linked data for. The entities we have linked for now are listed at http://www.opencalais.com/documentation/linked-data-entities. We will be adding more soon.
Thanks.
sumit