A Naive Bayes Classifier for NLP.

NBC_NLP.py

#The code for creating a Naive Bayes Classifier from text data stored in a pandas data frame

#Train Test Split The Data Frame
X_train, X_test, y_train, y_test = train_test_split(df.lemmatize_text, df.show_name, test_size=0.2, random_state=42)

#create a Scikit-Learn pipeline for Naive Bayes Classification
text_clf = Pipeline([('count_vectorizer', CountVectorizer()), 
                     ('tfidf_vectorizer', TfidfTransformer()),
                     ('clf', MultinomialNB())
                    ])

#Fit the training datat
text_clf.fit(X_train, y_train)

#Predict the categories of the test data
test_predictions = text_clf.predict(X_test)

#Evaluate the predictions based on the scripts' actual classes
print(metrics.classification_report(y_test, test_predictions, 
                              target_names = le.classes_))