The project is aimed at solving the problem of finding the sentiment of the user reviews on yelp.com
Libraries needed to run the code:
pip3 install -r requriements.txt
Download the Yelp dataset:
- Support Vector Machine
- Naive Bayes
- RCNN model
Data preprocessing
- Remove stop words
- Lemmetization
- Uppercase to Lowercase
- Removing any bad characters like \n, \t, $ etc using regex.
Check my blog on text preprocessing.
Due to an imbalance classes, F1 score was metric was used.