0100,0100,0100{ SEQ CHAPTER \h \r 1}In December I asked the same question about German Taggers that 0000,0000,8000Veronika Koller asked about today. I admit to my disgrace that I did not get around to sharing the information as promised (end of the semester, Christmas, Olympics, etc.) so now I have a motivation to do so. I’ll just cut and paste (and edit) the responses. Randall Jones -------------------- An earlier version of QTAG used to have a resource file for German - perhaps the author (Oliver Mason) could give you an upadated pointer for that 0000,8000,0000http://www.english.bham.ac.uk/staff/oliver/software/tagger/ Dr Tony Berber Sardinha LAEL, PUC/SP (Catholic University of Sao Paulo, Brazil) 0000,8000,0000tony4@uol.com.br 0100,0100,0100{ HYPERLINK "http://www.tonyberber.f2s.com" }0000,0000,FF00www.tonyberber.f2s.com ++++++++++++++++ There is one available at our institute. It is based on probabilistic methods (a HMM-model). You can have a look at it on 0000,8000,0000http://santana.uni-muenster.de/XlexPublic/ You might contact Prof. Dr. Wolf Paprotte (0000,8000,0000wolfp@marley.uni-muenster.de) to get more information. Sincerely yours Petra Steiner Arbeitsbereich Linguistik Universitaet Muenster Huefferstrasse 27 48149 Muenster Tel: 0251 / 83 39442 0000,8000,0000petra@marley.uni-muenster.de 0100,0100,0100{ HYPERLINK "http://santana.uni-muenster.de/~petra/" }0000,0000,FF00http://santana.uni-muenster.de/~petra/ ++++++++++++++++ You can try the German tagger from Zurich online at 0000,8000,0000http://www.ifi.unizh.ch/CL/tagger/ Regards, Gerold Schneider Department of Information Technology Computational Linguistics University of Zürich Winterthurerstr. 190 CH-8057 Zürich (Reply on this tagger also came from Oliver Mason) ++++++++++++++++ I would be happy to provide TnT. Please have a look at 0000,8000,0000http://www.coli.uni-sb.de/~thorsten/tnt/ TnT is trainable on different languages and tagsets. It comes with a pre-compiled language model for German (among others). -Thorsten Brants (Replies on this also from Andrew Bredenkamp, Caren Brinckmann, 0000,0000,8000Serge Sharoff, 0000,0000,8000Sandra Kuebler, 0000,0000,8000Brigitte Krenn, 0000,0000,8000Chris Brew) ++++++++++++++++ Conexor (www.conexor.fi) has an online demo of a German tagger. Regards, Atro Voutilainen If you provide us with a tokenlist one token (wordform or punctuation mark) per line. We can tag it for you with our morphological analysis system DMM, that provides not only a tag but also a complete morphosyntactic category. An alternative would be the Tree-Tagger by Helmut Schmid which assigns tags of the Stuttgart-Tübingen-Tagset. Markus Schulze Abteilung für Computerlinguistik Dipl.-Inf. Markus Schulze Bismarckstr. 6 Telefon: +49-9131-85-29252 91054 Erlangen Telefax: +49-9131-85-29251 0000,8000,0000http://www.linguistik.uni-erlangen.de/schulze ++++++++++++++++ Have a look at 0000,8000,0000http://www.conexoroy.com/products.htm Their tools aren't for free but quite robust. Hans Martin Lehmann Englisches Seminar 0100,0100,0100{ HYPERLINK "mailto:hmlehman@es.unizh.ch" }0000,0000,FF00hmlehman@es.unizh.ch0000,8000,0000 Universitaet Zurich Plattenstrasse 47 CH-8032 Zurich/Switzerland Phone: (41 1) 634 39 37 Home: (41 1) 462 18 51 ++++++++++++++++ You might want to have a look at the TreeTagger and the TnT-Tagger 0000,8000,0000http://www.ims.uni-stuttgart.de/projekte/corplex/TreeTagger/DecisionTreeTagger.html Best, Helmut Schmid Institute for Computational Linguistics (IMS-CL) University of Stuttgart 0100,0100,0100{ HYPERLINK "http://www.ims.uni-stuttgart.de/~schmid" }0000,0000,FF00http://www.ims.uni-stuttgart.de/~schmid (Also from Tanja Gaustad, Milena Slavcheva) ++++++++++++++++ I happen to be the author of a tagger considered by those which have exercised it as the best for German . It is currently licensed in Germany for nearly two years . On top of it I have built a parser which , judging by the statements of my Western interlocutors, would be at the same level . Coincidently it is currently in the process of licensing at the University of Zuerich , just mentioned by you ( Prof. Michael Hess ) . If you are interested in an academic license below the price of the market , I can send a demo for Linux . The parser is statistic , i.e. easily adaptable to other languages - at present there is also a version for English . Sincerely yours , Dr. ing. Vlad Gojol Senior Research Engineer Institutul National de Informatica Campul cu flori 6 , D25 , AP7 Bucuresti 77408 , Romania Phone : +40 1 778 5315 Email : 0000,8000,0000gojol@sunu.rnc.ro +++++++++++++++++ Besides the Brill tagger you mention, here are the ones I know about for German: 0000,8000,0000http://www.ims.uni-stuttgart.de/projekte/corplex/TreeTagger/DecisionTreeTagger- de.html 0000,8000,0000http://www.conexoroy.com/products.htmCourier New 0000,8000,0000http://www.coli.uni-sb.de/~thorsten/tnt/ 0000,8000,0000ftp://ftp.cis.upenn.edu/pub/adwait/jmx/ http://nats-www.informatik.uni- hamburg.de/~ingo/icopost/ I'm not sure whether these have been implemented for (or trained on or adapted for) German, but it should be possible: http://www.english.bham.ac.uk/staff/oliver/software/t agger/ 0000,8000,0000http://cs.nyu.edu./cs/projects/proteus/app/ 0000,8000,0000http://www.ltg.ed.ac.uk/software/ 0000,8000,0000http://www.scs.leeds.ac.uk/amalgam/amalgam/amalghome. htm 0100,0100,0100Times New Roman{ HYPERLINK "http://www.cis.upenn.edu/~mickeyc/stag/supertags.html" }0000,0000,FF00Courier Newhttp://www.cis.upenn.edu/~mickeyc/stag/supertags.html0000,8000,0000 0000,0000,8000Times New RomanDeryle Lonsdale0000,8000,0000Courier New Randall L. Jones Department of Germanic & Slavic Languages Brigham Young University