(Apologies for cross-postings)
The first public release of the RASP (robust accurate statistical
parsing) system in 2002 has been downloaded by over 120 sites and used
in diverse NLP and IR tasks.
The new release, which is also free for all non-commercial use, is
designed to address several weaknesses of the extant toolkit.
Go to:
http://www.informatics.susx.ac.uk/research/nlp/rasp/
for more information, licence and download details
Brief highlights:
1) All modules have been incrementally improved to cover a greater
range of text types.
2) The part-of-speech tagger lexicon has been semi-automatically
enhanced to better deal with rare or unseen behaviour of known words.
3) New facilities have been provided for user customisation of grammar
and output.
4) The grammatical relations output has been redesigned to better
support further processing.
5) Weighted grammatical relations can now be efficiently extracted
directly from the parse forest.
6) The training and tuning of the parse ranking model has been
made more flexible and accurate.
7) The entire pipeline has been modified to support XML input and
output.
This archive was generated by hypermail 2b29 : Fri Sep 22 2006 - 15:18:35 MET DST