This HTML version is for information purposes only. Please print the postscript version or the PDF version.
This agreement is made and entered into as of by and between the parties:
TnT, the short form of Trigrams'n'Tags, is a very efficient statistical part-of-speech tagger that is trainable on different languages and virtually any tagset. The component for parameter generation trains on tagged corpora. The system incorporates several methods of smoothing and of handling unknown words.
TnT consists of four modules: parameter generation ( tnt-para), the tagger ( tnt), comparison of original and tagged files ( tnt-diff), and token/type/tag counting ( tnt-wc). Two pre-compiled language models are supplied, one for German and one for English.
The work on TnT was made possible by support of Hans Uszkoreit at the department of Computational Linguistics, Saarland University, and by a grant from the Deutsche Forschungsgemeinschaft in the Graduiertenkolleg Kognitionswissenschaft Saarbrücken.
The German language model that is delivered with TnT is generated from the NEGRA corpus. This corpus was part-of-speech tagged and manually corrected at the Saarland University, Computational Linguistics, Saarbrücken, and University Stuttgart, Institut für maschinelle Sprachverarbeitung. It was also syntactically (structurally) annotated in Saarbrücken.
The English language model that is delivered with TnT is generated from the Susanne corpus. This corpus was published by Geoffrey Sampson and is described in his book English for the Computer, Clarendon Press, Oxford, 1995.
Copyright of TnT is to Thorsten Brants.
LICENSER grants USER a non-exclusive license to use TnT, version 1.0. USER agrees
Use of TnT or products derived from TnT for any commercial purposes requires explicit written agreement of LICENSER.
TnT will be held in confidence by USER and will not be disclosed by USER to third parties.
USER shall and will employ all necessary precautions to ensure that no persons or institutions other than persons as are in the employ of USER or in the same research project as USER will get access to TnT or parts thereof. Other persons or institutions desiring access to TnT should be directed to LICENSER to obtain separate licence agreements.
This license is granted by LICENSER to USER free of charge.
TnT and its documentation is provided on an ``as is'' basis, with no guarantee of its veracity or accuracy. No liability is accepted for any damage caused by its use.