The repo contains 3 ipython notebooks and 3 csv files and follwing is the way they are used
-
scraping .ipynb ipython notebook contains the code for scraping the reviews from the amazon website for 10 mobile phones. It uses the phone_list.csv for accessing the websites.
-
The output is the phone_review.csv file which contains the name of the model, the reviews and start ratings.
-
Training_DistilBERT.ipynb is the training of BERT model on the given data. The code was implemented on kaggle. It trains and saves the weights of the model.
-
testing_the_model.ipynb uses phone_review.csv and the trained weights. Creates a review_score.csv which contains the score predicted by the trained model.