Prediction of Customer Satisfaction in Airline Hospitality services for improved service delivery using Support Vector Machine

Abstract

The recent entry of Air peace airline service to London and competition resulting from this with other airlines showcased the trending competitiveness in airline industry which might affect the patronage of an airline over others. In line of this, ensuring customer’s satisfaction would enable business growth, sustainability and provision of improved service delivery. Gauging customer satisfaction would help in knowing steps to take to ensure better service delivery but employing traditional approach might be subjected to travel experience, time consumption, inaccurate which would hinder its objective. However, application of Machine learning techniques would give a faster, accurate and concise prediction resulting from historical customer’s reviews. This study aims to employing Support Vector Machine a traditional machine learning classifier for prediction of airline customer’s satisfaction in hospitality services employing an open access dataset comprising of different attributes like age, gender, flight distance among others. This dataset was pre-processed using various pre-processing techniques like normalization using standard scalar, transformation using label encoder, removal of missing values, feature selection using forward-backward feature selection techniques. The well pre-processed dataset was implemented using the SVM classifier with its three kernels for comparative analysis. The study used hold-out evaluation method with 80-20 split and gave average accuracies with Rbf, Linear and poly kernels of 93, 85 and 90% respectively. This study showed that the RBF kernel of SVM gave the best predictive performance with average accuracy, precision and F1-score of 93, 93 and 92% respectively with its good predictive airlines can identify the major factors that contribute to customer’s satisfaction to prioritize to improve service delivery and optimization of staffing at during strategic period. This would enhance customer’s satisfaction and ensure improved good service delivery.