Comments on: Part of Speech Tagging with NLTK – Part 3

By: Jacob Perkins

Jacob Perkins — Tue, 22 Jun 2010 21:05:32 +0000

I think you'll have to collect the stats manually. You could write a function like accuracy that takes in a “gold standard” of tagged sentences. Untag each sentence and run your tagger over it and compare it to the gold sentence. Count all the correct tags along with the total tags, then when it's finished you can calculate precision.

By: bullaggan

bullaggan — Tue, 22 Jun 2010 20:46:13 +0000

hi jacob,
how we can find precision ,recall and f,measure by using brill tagger,as brill's only displaying accuracy.and if we look at precision ,its formula is:

precision = Correctly tagged words by system
--------------------------------------------------- x 100%
Total no. of words tagged by taggers

so from where to get tagged words, that are Correctly tagged by nltk's brill.

i've successfully run brill tagger over Urdu language,nad prb is brill takes hexadecimal code ov urdu and also showing it in hexadecimal for urdu language.bt i need to know precision ,recall and f measure to quote in my FINAL Presentation, and my MS thesis plz...
i m short with time so help me out plzzzz.

By: Jacob Perkins

Jacob Perkins — Tue, 22 Jun 2010 14:05:32 +0000

By: bullaggan

bullaggan — Tue, 22 Jun 2010 13:46:13 +0000

By: Learning to do natural language processing with NLTK | JetLlib Journal

Learning to do natural language processing with NLTK | JetLlib Journal — Sun, 04 Apr 2010 22:12:55 +0000

[...] Part of Speech Tagging with NLTK – Part 1, Part 2, Part 3 [...]

By: Jacob Perkins

Jacob Perkins — Sun, 28 Mar 2010 21:58:22 +0000

Glad this helped, thanks for the positive feedback.

By: ke

ke — Fri, 26 Mar 2010 19:04:13 +0000

I'm working on a class project and this article series saved me a lot of time and trouble. It's much more accessible than the NLTK documentation, which I now only had to use to understand some specific details. Thanks a lot!

By: Jacob

Jacob — Wed, 16 Sep 2009 22:05:14 +0000

Cool! What’s your topic?
My name’s Jacob Perkins, and I should probably put it somewhere obvious

By: karen

karen — Wed, 16 Sep 2009 20:28:06 +0000

Hi! i’d like to cite this for my dissertation but i can’t find ur name anywhere!

By: Jacob

Jacob — Fri, 20 Mar 2009 17:00:03 +0000

I have not tested nltk.tag.pos_tag() (I’m pretty sure it wasn’t released when I wrote this series). I believe it was trained with most or all of the available corpora, which would definitely make it more accurate. However, it’ll only have high accuracy for text that’s similar to the corpora it was trained on. If you’re tagging text that has a lot of specialty/unique words and phrases, you’ll need to create your own training data for the training process in order to get accurate results.