This project is about creating a supervised NB classifier. This is a 2 step process:
- It creates a NB model from already tagged data.
- Use this model to tag unseen data.
In this project we are revieweing the reviews of hotels and classifying into 4 tags :
- Positive
- Negative
- Truthful
- Deceptive
We use Naive Bayes theorem to construct a NB Model. we build the model using a corpus which is already tagged by a human. From the tagged sentenses, we build the NB Classifier. For handling unseen words, we perform Laplace Smoothing by performing add 1 smoothing. The accuracy of this model is 0.90
This is written in Python