Airline Customer Reviews Analysis and Booking Completion Prediction using Data Visualization and Machine Learning for British Airways

This study focuses on analyzing airline customer reviews and predicting booking trends, specifically for British Airways. The objectives include interpreting customer sentiments from reviews, identifying key themes, and examining the relationship between these sentiments and factors like flight duration and service offerings. A significant aspect of this research is developing a predictive model to forecast booking outcomes. Methodologically, the study utilizes the Extra Trees Classifier for feature selection, focusing on 8 primary features, and employs the SMOTE technique for dealing with an unbalanced dataset. The predictive models experimented comprise of Random Forest, Logistic Regression, Naive Bayes, and K-Nearest Neighbors (KNN). The models achieved an accuracy rate of 96%, 94%, 92%, and 94% respectively for booking prediction. The results demonstrate the effectiveness of combining machine learning and text analytics in the airline industry, offering valuable insights for strategic decision-making.