MEETING NOTES

DONE

Chris did
  Cleaning out usage of io.StreamTokenizer (except in NegraExportReader)
  Making NBClassifier support having a prior

Roger did nothing

Galen got rid of parser global options
Got rid of constructors with length
 - but should add in setMaxLength method so that it's easy to do.

Bill: English grammatical role finder still in progress
 - Have made parser able to return NP-TMP trees
 - Have EnglishTreebankParserParams inherit from AbstractTreebank....
 - Made EnglishTest class

Dan did nothing

Chris Cox and Jamie Nicolson were at JavaNLP for the first time.

Other people were away/absent (Jenny, Kristina, Mona, Teg).

NEW TASKS

Roger should remove use of io.StreamTokenizer from NegraExportReader.

Galen wants to clean up TreeProcessor, TreeTransformer, TreeNormalizer.
 TreeMangler.  Chris or others can help?

Jenny and Chris Cox shoud do unified common front end processing.

Kristina will also be away next week, and it seems like we can't really
discuss new integration of classifiers until she is back, but she said
she would write up and distribute a document listing classifier
classes/packages, what classification problem and algorithm they
implement, what input data representations they support, and what
internal data representations they use, and we could then start to think
about how to extend and unify things.