Cool. FWIW, there is a repo (not updated in a while) for collecting models and related code and licenses:
https://github.com/utcompling/OpenNLP-Models I'd love it if someone were to take this, organize, and extend... :) -Jason On Thu, Oct 31, 2013 at 3:21 PM, Thomas Zastrow <[email protected]>wrote: > Dear all, > > As promised, here is now a first version of the "NE Model for German, > Politics" (NEMGP). > > It is based on 5094 sentences from Wikipedia and Wikinews. Its > performance is not perfect, but quite good in the domain for which it > was build for - texts around German politics. My plan is to extend it in > the future, I will also release the editor I build for doing NE > annotation. Somewhere in the future :-) > > Please feel free to play around with the models and the data - I put the > original training data, a binary model for OpenNLP and one for Stanford > NE Recognizer online. > > Its now under a Creative Commons License, which should be compatible > with the CC license of the original Wikipedia texts. > > Here is the link: > > http://www.thomas-zastrow.de/nlp/ > > Best, > > Tom > > P.S: Its just a hobby project. > -- Jason Baldridge Associate Professor, Dept. of Linguistics, UT Austin Co-founder & Chief Scientist, People Pattern http://www.jasonbaldridge.com http://twitter.com/jasonbaldridge
