Institut

Studium

Forschung


 

ims_german_POS.scm next up previous contents
Next: ims_german_debugtools.scm Up: Files in festival/lib/german Previous: Files in festival/lib/german &nbsp Contents

ims_german_POS.scm

This file deals with POS tagging for German. Two different tagging methods are available:
1. No_POS that simply does nothing at all
2. The Treetagger by Helmut Schmid, which is an external package not (yet) included in the opensource version. In this file all parameters are set that are required for the treetagger.
In TTS two tagsets are used as defined by the parameter 'pos_tagset':
1. stts (STuttgart TagSet /fak5/ims/ftp/pub/corpora/stts_guide.ps.gz
2. ustts (``underspecified`` STTS tagset) ustts contains the 11 main categories of STTS ADJ, ART, ADV, KO, ITJ, N, CARD, AP, P, PTK, V
other tagsets as provided by several lexicons (e.g. bomp, celex) are currently mapped to ustts using a table defined in this file (in the post_hooks of the lexicon)



Martin Barbisch
2001-08-28