Through this project I will be building a newspaper classifier for five categories: business, entertainment, politics, sports and tech. To do this, I will be taking different methods including 1) bag of words approach using TF-IDF, 2) neural network (CNN and/or LSTM) and 3) neural network using 1) and document summary as features. In the end, I will evaluate the different methods using test data. My hypothesis is that method 3) will bring the best result.
