Hotel-Reviews-Classification

How genuine are the reviews? Are they positive or negative?
Find out with this classifier.
It's a Naive Bayes classifier that classifies hotel reviews in class Fake/True and Positive/Negative.

Training file format: File train-labeled.txt containing labeled training data with a single training instance (hotel review) per line. The first 3 tokens in each line are:

1. a unique 7-character alphanumeric identifier
2. a label True or Fake
3. a label Pos or Neg

These are followed by the text of the review.

Results

Reference Mean F1	My Mean F1
0.8937280481685901	0.9047

Future scope : It can be extended to classify any types of reviews.

For more details refer : http://ron.artstein.org/csci544-2018/coding-2.html and https://github.com/visrahane/Hotel-Reviews-Classification/blob/master/Description.pdf