This repo is for final project of CSE256 NLP at UCSD. The core task is to use binary logistic regression classifier to complete text classification. It contains two task:
- Task1: sentiment classfication based on reviews
- Task2: text classification based on clinical statements
We also leveraged Lime and Shap to explain the models with visualization.
Jun 3th 2019
- Form Validation: no more than 2 words
- Lime highlight for task 1
- Shap Force bar for task 2
- adaptive height for shap force bar