Sentiment-Analysis

Goal

The project is aimed at solving the problem of finding the sentiment of the user reviews on yelp.com

Installation

Libraries needed to run the code:

pip3 install -r requriements.txt

Download the Yelp dataset:

Algorithms

1. Support Vector Machine
2. Naive Bayes
3. RCNN model

Steps involved

Data preprocessing

1. Remove stop words
2. Lemmetization
3. Uppercase to Lowercase
4. Removing any bad characters like \n, \t, $ etc using regex.

Check my blog on text preprocessing.

Evaluation

Due to an imbalance classes, F1 score was metric was used.