CS 410 Final Project: Classification Competition

Preparation

Datasets used are test.jsonl and train.jsonl in the data folder.

Required Packages

• scikit learn
• numpy
• xgboost
• Pytorch
• matplotlib
• seaborn
• nltk
• json
• pickle
• pandas
• re
• math
• tensorflow
• sys
• string
• gensim
• os
• keras

Baseline Model

All the work done within the jupyter notebook Text Classification baseline.ipynb

CNN Model

All the work done within the jupyter notebook CNN-final.ipynb

BERT Model

All the work done within the jupyter notebook BERT.ipynb

Documentation

Documentation.docx

Video Illustration

https://youtu.be/UJcDlOYOAZY

Reference

CNN: https://towardsdatascience.com/cnn-sentiment-analysis-1d16b7c5a0e7

BERT: https://colab.research.google.com/drive/1pTuQhug6Dhl9XalKB0zUGf4FIdYFlpcX#scrollTo=DEfSbAA4QHas