This project dealt with implementing the classic Naive Bayes algorithm to detect spams in a dataset of emails. We used different kinds of classifiers like binary classifier, term frequecy classifier to test the performance of the algorithm. Smoothing techniques were used to further improve the performnce.