Well the additional context thing we never got right. Its not really
supported
during training with the new tools and makes using the name finder a bit
more difficult
because the additional context needs to be passed in on every call.
For testing/research I suggest to simply include a POSModel in the name
finder
model and implement a new feature generator which can use it to do pos
tagging
on the passed-in sentence.
Would that work for you?
And that is exactly for what the resources are there for, with the
resources your custom
POSFeatureGenerator could access the POSModel resource.
Jörn
On 04/02/2012 02:35 PM, Svetoslav Marinov wrote:
Hi all,
Is there any possibility to use additionalContext with the NameFinderME.train?
If so, how? If there isn't maybe this should be an issue to be added in the
future releases?
I would REALLY greatly appreciate if someone can help (give me some sample
code/show me) how to add POS tag features while training and testing
NameFinder.
If the incoming data is just tokens with NO POS tag information, where is the
information taken then? A new file? Run a POS tagging model before training? Or?
And what is the purpose of the resources (i.e.
Collection.<String,Object>emptyMap()) in the NameFinderME.train method? What
should be ideally included in there?
I just can't get these things from the Java doc API.
Thanks in advance!
Best,
Svetoslav
