Code written for the dissertation project during my MSc in Business Analytics at the University of Edinburgh: Text classification of hotels marketing emails
The code leverages NLP analysis. It focuses on tonality classification, and sentiment-analysis algorithms to detect which writing style created the most engagement from the customers.
Please note that any reference to the hotels data was deleted for confidentiality issues. It has therefore become impossible to run the final notebook making predictions on the hotels e-mails.
The straightforward way to run the double-classifier model used to be by running the PREDICTION_HOTELS_EMAILS notebook.