OpenNLP NE tagger German Model

Hello,

currently I'm working on a project where I try to enrich my text with POS tags. (Especially Person/Location/Organization)

The OpenNLP NE tagger works fine for english texts. In my case the text is german and for that I'm using the model provided here: http://opennlp.sourceforge.net/models-1.5/.

Both German Models (Maxent and Perceptron) result in Errors.

WARN 	 OpenNlpModel                  	 Maxent model could not be loeded from file [<FILEPATH>\de-pos-maxent.bin]!
ERROR	 TaggerCellFactory             	 Tagger could not be borrowed from pool.

As a workaround I'm using the Dictionary tagger but I hope that there is a better solution.

Thanks in Advance,

Thomas

 

Hi Thomas,

the OpenNLP NE Tagger can only handle "Name Finder" Models from http://opennlp.sourceforge.net/models-1.5/

The model you want to use "de-pos-maxent.bin" is a POS tagger model. If you want to do POS tagging on German texts please use the node "Stanford tagger" which is using the Stanford POS tagging lib. The node provides three models for German POS tagging, that can be selected in the node dialog. The model "German fast" ist the fastest.

Cheers, Kilian

Thank you for the fast answer.

I already used the Stanford tagger. But in my specific case I need to differentiate between Persons and Locations.

OpenNLP doesn't seem to offer Name Finder models for ther german language.

Is there anyway besides the Dictionary tagger to get NER for Persons/Locations for the German language in KNIME?

Best Regards

 

Hi phyte,

I am afraid the only two ways to tag persons / locations for German language is

a.) use the dictionary tagger

b.) train a Name Finder model using the OpenNLP and use that model with the OpenNLP Node in KNIME

http://stackoverflow.com/questions/11204352/training-own-model-in-opennlp

http://opennlp.apache.org/documentation/1.5.3/manual/opennlp.html

or, well there is a third option

c.) implement a KNIME tagger node using another tagging lib

http://tech.knime.org/for-developers-integration-of-custom-tagger

http://tech.knime.org/for-developers-integration-of-custom-tag-sets

Cheers, Kilian