Project Link : https://github.com/piyushP7pravin/Text-Classification

MultinomialNB implements the naive Bayes algorithm for multinomially distributed data, and is one of the two classic naive Bayes variants used in text classification.

I have used Multinomial Naive Bayes algorithm to classify the given text document to different categories.
I have used 20,000 documents from 20 newspaper categories, made dataset from the given documents. Built a Multinomial Naïve Bayes Classifier from Scratch that classifies any test document to one of the 20 categories. Made a comparison report between the results of our Naive Bayes implementation with Sklearn Multinomial Naïve Bayes Classifier.
Overall accuracy seems to be almost same for both cases.