I am using GATE for Arabic under Linux and when using Stanford Tagger there aren't any POS tags. All parameters were not changed (as default) what can I do to perform that task? Does this problem depend on the parameters ? What installation directory of the tagger must be set in the taggerBinary parameter?
Asked
Active
Viewed 252 times
1 Answers
1
First you have to create a new StanfordPOSTaggerPR in GATE and initialize the Tagger with the arabic.tagger
model provided with the StanfordTagger.
The last version of GATE is the 8.0 and it uses the StanfordTagger 3.4. So you'll have to download the models provided with this version.
Next, you have to create a corpus pipeline with a SentenceSplitter and a Tokeniser (I've tried with the UnicodeTokeniser
and the RegExp SentenceSplitter
):
Finally, try the pipeline with a sample file:

Chester Mc Allister
- 437
- 3
- 19