Naive-Bayes-Classifier

Using Naive Bayes Classifier to distinguish article of different author or different language

Bernoulli Naive Bayes Classfier

Using Bernoulli Naive Bayes Classfier to distinguish article of different author, using stop words.
Input:

• Train and test set
• stop words
• ground truth table

Output:

• Accuarcy
• Confusion Matrix
• Feature Ranking of top 20 by their class-conditional entropy
• Feature Curve

Run like this: python3 Bernoulli_Naive_Bayes.py problemA/

Multinomial

Using Multinomial Naive Bayes Classfier to distinguish article of different language. Using both bigram and trigram.
Input:

• Train and test set

Output:

• Accuarcy
• Confusion Matrix

Run like this: python3 Multinomial_Naive_Bayes.py extra-test/

NOTE:The dataset we use is a subset of the Ad-hoc Authorship Attribution Competition(AAAC) Dataset, publicly available at http://www.mathcs.duq.edu/~juola/authorship_materials2.html.