Hello,
on which data did you train the models? If the data is publicly
available it would be better if you could add
format support for it to OpenNLP.
Anyway, we are also happy about model contributions. To be able to
distribute the models they need
to be licensed under AL 2.0.
Jörn
On 05/30/2014 10:14 AM, Slawomir Krol wrote:
Dear all,
I've been working on Polish language support for some time now and
I've got some Maxent models to share (sentence detection, tokenizer,
POS tagger, more to come).
First, is it fully ok with you to add binaries for Polish to your
models' download page?
Second, my boss would be particularly pleased if there was some way to
acknowledge my company as the contributor of the binaries (say, put
"trained on Polish National Corpus data @ IBM" in the description).
I'm looking forward to hearing from you.
Sincerely yours,
*Sławomir Krystian Król*
Software Engineer
Business Analytics
IBM SWG, Krakow, Poland
Phone: +48 500 72 82 77
e-mail: [email protected]
IBM Polska Sp. z o.o. oddzial w Krakowie, ul. Armii Krajowej 18,
30-150 Krakow, NIP: 526-030-07-24
Sad Rejonowy dla m.st. Warszawy, XII Wydzial Gospodarczy KRS KRS
0000012941, Kapital Zakladowy: 42.153.600 PLN
