Naiive-Bayes-Classifier-on-Yelp-Reviews

A Naiive Bayes Classifier that takes in Yelp reviews as input to determine whether the reviews are positive or negative.

Pre-processing step

This step converts each sentence in into a feature vector (one-hot encoding) and each sentence are represented as a "bag of words".

1. strip the punctuation
2. convert train and test set into one hot encoding
3. output the preprocessed data into two saperate files

To Run

In the terminal do python bayes.py

Training accuracy: 0.9679 Test accuracy: 0.7746