Contd: Zemanta API: from text to LOD

Kingsley Idehen wrote:
>
> Andraz Tori wrote:
>> Hi,
>>
>> I am Andraz Tori, CTO at Zemanta. This is my first post to this list.
>> I've read the discussion about commercial announcements just a day ago,
>> so I hope this this announcement is relevant enough albeit it comes from
>> a for-profit company.
>>   
> Welcome!
>
> Relevant!
>
> Please feel at home :-)
>> I'd like to announce that Zemanta today launched a semantic API that is
>> able to take plain text and disambiguate most important entities found
>> to their meanings in Linking Open Data (currently dbPedia, MusicBrainz,
>> Semantic CrunchBase and Freebase).
>>   
> I can't find the API itself or API demos.
>
> I've seen your illustrations etc..
>> As far as we know this is first such undertaking of this scale.
>>   
> Not necessarily, but being first doesn't matter, it's the value 
> delivered that is ultimately the key factor (imho) :-)
>> API is free to use up to 10k calls per day (default limit is 1k, email
>> needs to be sent to raise it). Response can be in XML, JSON or RDF/XML.
>>
>> --------------------------------------------------------------------
>>
>> Developer info can be found at http://developer.zemanta.com and less
>> technical explanations at http://www.zemanta.com/api
>>
>> I am very interested if this will make any new types of Mashups happen
>>   
> I would say potentially interesting meshups with the output of other 
> Linked Data aware services, for instance. Everything meshes with 
> everything in this realm once the essence is understood.
>> Also I am interested in feedback on RDF/XML format/ontology.
>>   I see ourselves a bit like a stargate portal from unstructured text 
>> into
>> the LOD, enabling LOD to be used in broader number of situations.
>>
>> Comments welcome :)
>>
>>   
> Please send a link to a page that provides sample Linked Data example.
>
> Also clarify if you offer a REST style service as I only see your 
> language specific APIs in your document web collective re. APIs etc..
>
> I am assuming you offer something that works like this:
>
> 1. You receive a chunk of blurb as input
> 2. You analyze (where extraction and disambiguation occurs)
> 3. Return an RDF Linked Data graph in one or more formats as part of 
> you output
>
>
>
Andraz,

I found the following which address some of my initial questions:

1. http://developer.zemanta.com/docs/suggest/
2. http://developer.zemanta.com/docs/suggest_markup/


So I am set for now :-)

-- 


Regards,

Kingsley Idehen       Weblog: http://www.openlinksw.com/blog/~kidehen
President & CEO 
OpenLink Software     Web: http://www.openlinksw.com

Received on Tuesday, 9 December 2008 19:38:25 UTC