Chapter 6, exercise 5

## -*- coding: utf-8 -*-
import nltk
import random

def gender_features(word):
    return {'last_letter': word[-1],
            'last_two_letters': word[-2:],
            'last_three_letters': word[-3:]}

maleNames = [unicode(name, 'latin-1')
             for name in nltk.corpus.names.words('male.txt')]
femaleNames = [unicode(name, 'latin-1')
               for name in nltk.corpus.names.words('female.txt')]

names = [(name, 'male') for name in maleNames] + \
        [(name, 'female') for name in femaleNames]
random.shuffle(names)

featuresets = [(gender_features(n), g) for (n,g) in names]
train_set, test_set = featuresets[500:], featuresets[:500]

naiveBayesClassifier = nltk.NaiveBayesClassifier.train(train_set)
decisionTreeClassifier = nltk.DecisionTreeClassifier.train(train_set)
maxEntClassifier = nltk.MaxentClassifier.train(train_set)

print 'Accuracy with the naive Bayes classifier : {0:.2f}%'.format(nltk.classify.accuracy(naiveBayesClassifier, test_set))
print 'Accuracy with the decision tree classifier : {0:.2f}%'.format(nltk.classify.accuracy(decisionTreeClassifier, test_set))
print 'Accuracy with the maximum entropy classifier : {0:.2f}%'.format(nltk.classify.accuracy(maxEntClassifier, test_set))