[Top] [All Lists]

Re: [ontolog-forum] Semantic Web & NLP Technology Combo: TellMeFirst

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: Kingsley Idehen <kidehen@xxxxxxxxxxxxxx>
Date: Mon, 23 Sep 2013 15:11:22 -0400
Message-id: <5240925A.40106@xxxxxxxxxxxxxx>
On 9/23/13 1:56 PM, John F Sowa wrote:
On 9/23/2013 12:23 PM, Kingsley Idehen wrote:
yet another example of a contemporary service that leverages tools
based on the Semantic Web technology stack.

TellMeFirst (http://tellmefirst.polito.it ) is a tool for classifying
and enriching textual documents via Linked Open Data.
Thanks for the pointer.  It's an interesting demo.

But I find the simple summaries that Google generates from their
"knowledge graph" more useful.  They're also less annoying than
the glitzy graphics that make an impressive demo.

I'll admit that DBpedia can be useful as one component of a more
complete system, as it was for Watson.  A good system doesn't
need all the power of Watson, but it needs some way (eg, click
patterns) to evaluate and rank relevance.

In summary, DBpedia is useful as a component of a larger system.
But a helpful system that can enrich a document needs (a) much more
detailed analysis of the documents, (b) methods for determining what
kinds of information are relevant to the text, and (c) methods for
to evaluate and rank relevance.

No doubt about that. The ecosystem around the likes of DBpedia is growing rapidly across many dimensions, including those you mention above.

Above all, a good interface must *not* be annoying.  Does anybody
remember Microsoft's Clippy?

The beauty of Linked Data is that you are not confined a an specific presentation interface, as along as the URIs that denote entities adhere to Linked Data principles. Here's are some examples of what I mean:

1. http://t.co/dXcsh3MObl -- Query Definition (inside PivotViewer hosted SPARQL Query Editor)

2. http://t.co/EF4fGCmHoo -- Pivot Viewer Report (you might need to click "edit" then "view" due to issue with permalink and state)

3. http://t.co/jONBiyHIgI -- SPARQL Query Results Page (just follow-your-nose by clicking on links of interest for full Linked Data experience)

4. <http://edostate.rww.io/data/foodcenters#152> -- sample URI from the results in #3 (i.e., just click one of the items in the results table) that resolves to yet another UI

5. http://t.co/OTVE3aBv1A -- SPARQL Query Results in CSV format

6. http://t.co/ur1mFGgsfn -- Google Spreadsheet Import of Data from SPARQL CSV output URL .




Kingsley Idehen	      
Founder & CEO 
OpenLink Software     
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)

<Prev in Thread] Current Thread [Next in Thread>