Practice Competition on Kaggle
Competition Website
Individual Practice Work
Classify the sentiment of sentences from the Rotten Tomatoes dataset
- get-glove.sh downloads glove vectors from stanford.edu
- requirements.sh fulfils package requirements
- glove.py loads glove vectors (found it on standford.edu and adapted it for my needs)
- models.py contains model definitions
- visualize.py plots visualizations
- utils.py stores utility functions
- sentiment.py is the main program
- nltk
- numpy
- matplotlib
- keras
- scikit-learn
- pandas
- graphviz
- pydot
If conda is installed, run requirements.sh script and supply python3 environment name.
bash requirements.sh
- Clone the repository
git clone https://github.com/pawanmsr/sentiment-analysis-on-movie-reviews.git - Install all the necessary packages mentioned in Requirements
- Download the kaggle dataset from Competition Website and extract it inside the cloned repository
- Run sentiment.py
python sentiment.py ARGUMENTSEnter python sentiment.py -h for help
TODO: describe various models used