Python Code to train a Hidden Markov Model, using NLTK

hmm-example.py

__author__ = 'ssbushi'

# Import the toolkit and tags
import nltk
from nltk.corpus import treebank

# Train data - pretagged
train_data = treebank.tagged_sents()[:3000]

print train_data[0]

# Import HMM module
from nltk.tag import hmm

# Setup a trainer with default(None) values
# And train with the data
trainer = hmm.HiddenMarkovModelTrainer()
tagger = trainer.train_supervised(train_data)

print tagger
# Prints the basic data about the tagger

print tagger.tag("Today is a good day .".split())

print tagger.tag("Joe met Joanne in Delhi .".split())

print tagger.tag("Chicago is the birthplace of Ginny".split())

"""
Output in order (Notice some tags are wrong :/):
[('Today', u'NN'), ('is', u'VBZ'), ('a', u'DT'), ('good', u'JJ'), ('day', u'NN'), ('.', u'.')]
[('Joe', u'NNP'), ('met', u'VBD'), ('Joanne', u'NNP'), ('in', u'IN'), ('Delhi', u'NNP'), ('.', u'NNP')]
[('Chicago', u'NNP'), ('is', u'VBZ'), ('the', u'DT'), ('birthplace', u'NNP'), ('of', u'NNP'), ('Ginny', u'NNP')]
"""

I've tried to run

tagger.evaluate(treebank.tagged_sents()[3000:])

result 0.36844377293330455

the code doesn't work properly for latest version of NLTK
you should create tagger so

tagger = nltk.HiddenMarkovModelTagger.train(train_data)

then

tagger.evaluate(treebank.tagged_sents()[3000:])

shows 0.8984243470753291

Yes, tagger = nltk.HiddenMarkovModelTagger.train(train_data) worked for me too.

Thanks for your updates, @achmedzhanov and @krisGTech. This is a very old snippet obviously, but your comments might help others find an short and easy example to get started with HMMs.