newsgroup.py

from sklearn.datasets import fetch_20newsgroups
from sklearn.feature_extraction.text import CountVectorizer
from sklearn.feature_extraction.text import TfidfTransformer
from sklearn.linear_model import SGDClassifier
from sklearn.pipeline import Pipeline
from sklearn import metrics

newsgroups_train = fetch_20newsgroups(subset='train')
newsgroups_test = fetch_20newsgroups(subset='test')

pipeline = Pipeline([
    ('vect', CountVectorizer(ngram_range=(1, 2), max_df=0.5)),
    ('tfidf', TfidfTransformer(sublinear_tf=True)),
    ('clf', SGDClassifier(loss='hinge', penalty='l2',
        alpha=0.0001, max_iter=50, tol=0.0005)),
])

print('training')
pipeline.fit(newsgroups_train.data, newsgroups_train.target)

predicted = pipeline.predict(newsgroups_test.data)
accuracy = metrics.accuracy_score(newsgroups_test.target, predicted)
print(accuracy) # 0.865

# Other things to explore via supervised bag-of-words method:
# - Truncated singular value decomposition and latent semantic analysis
# - CountVectorizer analyzer and n-grams