

In [ ]:

    
import pandas as pd



In [ ]:

    
data = pd.read_csv('/Users/walter_tyrna/Finalized_DS_Code/cc_title_may.csv', header = 0 , names = ['country','title'])

The csv file only contains titles and countries of origin



In [ ]:

    
data.shape

(77245, 2)



In [ ]:

    
data['country_number'] = data.country.map({'en-us':0, 'en-uk':1, 'en-au':2, 'en-in':3, 'en-ca':4, 'fr-fr':5})



In [ ]:

    
data.head(10)

country title country_number 0 en-au Indigenous Man Stuck In The US After Sudden De... 2 1 en-us I Wore Pinterest-Style Updos For A Week And Th... 0 2 en-us The Mountain From "Game Of Thrones" Has A Ridi... 0 3 en-us What Does Your Favorite Milkshake Flavor Say A... 0 4 en-us How Well Do You Actually Know Your Best Friend? 0 5 en-us Men Recreate Iconic Photos And Get Photoshoppe... 0 6 en-us 14 Expert Ways To Tell If Clothes Are Well-Mad... 0 7 en-au 19 Times Chrissy Teigen Was Goddamn Hilarious ... 2 8 en-uk Only A Foodie Can Get More Than 70% On This Quiz 1 9 en-au 21 Tumblr Posts About Sexting Guaranteed To Ma... 2



In [ ]:

    
X = data.title
y = data.country_number



In [ ]:

    
from sklearn.cross_validation import train_test_split
from sklearn.feature_extraction.text import CountVectorizer



In [ ]:

    
X_train, X_test, y_train, y_test = train_test_split(X, y, random_state = 1)



In [ ]:

    
vect = CountVectorizer()



In [ ]:

    
X_train_dtm = vect.fit_transform(X_train)



In [ ]:

    
X_test_dtm = vect.transform(X_test)



In [ ]:

    
from sklearn.naive_bayes import MultinomialNB



In [ ]:

    
nb = MultinomialNB()



In [ ]:

    
nb.fit(X_train_dtm, y_train)

MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)



In [ ]:

    
y_pred_class = nb.predict(X_test_dtm)



In [ ]:

    
from sklearn import metrics



In [ ]:

    
metrics.accuracy_score(y_test, y_pred_class)

0.9474420049710025



In [ ]:

    
metrics.confusion_matrix(y_test, y_pred_class)

array([[11115, 262, 216, 78, 32, 13], [ 186, 3569, 34, 11, 8, 1], [ 49, 52, 2229, 6, 6, 1], [ 35, 6, 0, 958, 0, 0], [ 11, 7, 1, 0, 426, 0], [ 0, 0, 0, 0, 0, 0]])



In [ ]:

    
from sklearn.pipeline import Pipeline



In [ ]:

    
model = Pipeline([
    ("vectorizer", CountVectorizer()),
    ("classifier", MultinomialNB()), 
    ])



In [ ]:

    
test = ["17 Mean Girls Quotes That Sum Up Life In Australia", "21 Fucking Funny Aussie Olympic Tweets",]



In [ ]:

    
model.predict(test)

array([0, 2])