Predicting Customer Loyalty in the Airline Industry: A Machine Learning Approach Integrating Sentiment Analysis and User Experience

Abstract

Ensuring customer loyalty is crucial for the success of any airline service provider in today's competitive environment. This study employs machine learning techniques to predict the likelihood of customers revisiting airline services, emphasizing the role of emotional connections in fostering loyalty. By analyzing feedback comments and satisfaction ratings, we explore how sentiments expressed by customers correlate with their propensity to return. Using sentiment analysis and features extracted through the Linguistic Inquiry and Word Count (LIWC) methodology, we categorize sentiments into various dimensions, integrating these with user experience (UX) elements for a comprehensive predictive model. Our methodology includes a robust data collection process, involving an initial survey of 17,000 valid responses and a follow-up survey one year later. We evaluate multiple classifiers, including Decision Tree, Random Forest, and XGBoost, through five-fold cross-validation. Results reveal that XGBoost achieves the highest accuracy of 85% in predicting return visits, highlighting the predictive power of machine learning in understanding customer behavior. These findings offer significant insights for airlines, enabling them to tailor services and strategies to enhance customer satisfaction and loyalty. Our study underscores the importance of sentiment analysis and UX in predicting customer loyalty, providing a roadmap for future research and practical applications in the airline industry.