[okfn-labs] Named entities (was: Find country names in blobs of unknown text)

Fri Jun 20 09:00:06 UTC 2014

Hi Friedrich,
it depends on what are your needs probably.

But there is a difference on the outputs: if you want to extract an entity
from a text, you might want also to disambiguate and having a correct
context...
And having also a related source of information, to enrich the text too.

So, talking about named entity extraction, citing from Wikipedia[1]:
"A recently emerging task of identifying "important expressions" in text and
cross-linking them to Wikipedia [21] [22][23] can be seen as an instance of
extremely fine-grained named entity recognition, where the types are the
actual Wikipedia pages describing the (potentially ambiguous) concepts."

This is called also Entity Linking.[2]
This is what you can do with dataTXT API, for example.

>setting up targeted NER services that would run against a pre-defined,
restricted set of entities?

Can you give some examples/samples texts? 
I mean: with dataTXT we have the custom spot featured, as I mentioned
before, to match some text labels against some entities which already exist
in Wikipedia/DBpedia. 

The other thing you can do is personalize/extend dandelion.eu, the knowledge
graph underlying dataTXT: this feature will be available during this summer.
Ah, we are looking for beta testers too! .)

 Matteo
 --
 @dagoneye

[1] - http://en.wikipedia.org/wiki/Named-entity_recognition
[2] - http://en.wikipedia.org/wiki/Entity_linking

--
View this message in context: http://okfn-labs.28008.n7.nabble.com/okfn-labs-Find-country-names-in-blobs-of-unknown-text-tp33p53.html
Sent from the okfn-labs mailing list archive at Nabble.com.