A colab notebook in R containing sentiment analysis on US Airline tweets.
This project was developed as a part of the course Data Mining, Text Mining and Big Data Analytics - Alma Mater Studiorum Università di Bologna.
The notebook provides an in-depth sentiment analysis of tweets related to US Airline companies in 2015. In the first section the analysis is performed using the syuzhet package for sentiment scores and emotion classification, in the second part machine learning models (Gradient boosting decision trees, Naive Bayes) are trained to predict the sentiment of the tweets. Finally, the models are compared with statistical methods on the basis of their performances on the test set.