Corpora: POS tagger and collocation statistics

Michael Nelson (mnelson@ra.abo.fi)
Sun, 4 Apr 1999 16:04:24 +0300 (EET DST)

Dear Colleagues,

Can anyone help me with the following two queries:

1. Is there a POS tagger available that works under Windows that is a)
resonably reliable and b) doesn't cost the earth? All the taggers I've heard
about work only under Unix or Linux and I don't have access to either.

2. Is there a program available, again under Windows, that can give me t,z
and MI scores for collocational significance?

At present I am using Mike Scott's marvellous WordSmith and this does give
me MI scores. But MI of course tends to throw up collocates of very low
frequency and I'd like to be able to compare the 3 different statistical
measures to see what the differences are - I'm working on a 1 million word
corpus I've created for my PhD.

Any help greatly appreciated,

Mike Nelson