Simultaneously Classifying News Categories and Extract Keywords This project is bulit to learn classifiers to automatically classify news categories give their text and title. Three methods have been tested:
- Logistic Regression
- LSTM
- Transformer
The confusion matrix of test set is shown below:
Further, we have extracted the important words for each category using the weights of Logistic Regression's weights, which is depicted in the below figure.
