Sentiment Analysis - Model Creation

pickling_classifier.py

import nltk
import pickle
from nltk.classify.scikitlearn import SklearnClassifier

def pickling(file, document_name):
	save_documemts = open('../pickled_algos/' + document_name + '.pickle', 'wb')
	pickle.dump(file, save_documemts)
	save_documemts.close()

# let the training_set contains data in the form of -
# (sentence, sentiment), e.g.
# This is a nice tea, 1
# This is a bad tea, 0
training_set

# Original Naive Bayes
classifier = nltk.NaiveBayesClassifier.train(training_set)
pickling(classifier, 'original_naive_bayes_classifer')

# Multinomial Naive Bayes
MNB_classifier = SklearnClassifier(MultinomialNB())
MNB_classifier.train(training_set)
pickling(MNB_classifier, 'MNB_classifier')

# Bernouli Naive Bayes
Bernoulli_classifier = SklearnClassifier(BernoulliNB())
Bernoulli_classifier.train(training_set)
pickling(Bernoulli_classifier, 'Bernoulli_classifier')

# Logistic Regression Classifier
LogisticRegression_classifier = SklearnClassifier(LogisticRegression())
LogisticRegression_classifier.train(training_set)
pickling(LogisticRegression_classifier, 'LogisticRegression_classifier')

# Linear SVC Classifer
LinearSVC_classifier = SklearnClassifier(LinearSVC())
LinearSVC_classifier.train(training_set)
pickling(LinearSVC_classifier, 'LinearSVC_classifier')

# NuSVC Classigier
NuSVC_classifier = SklearnClassifier(NuSVC())
NuSVC_classifier.train(training_set)
pickling(NuSVC_classifier, 'NuSVC_classifier')

# SGDC Classifier
SGDC_classifier = SklearnClassifier(SGDClassifier())
SGDC_classifier.train(training_set)
pickling(SGDC_classifier, 'SGDC_classifier')