Enhancing Service Quality and Accessibility in Airports: Insights from Automated Social Media Analysis

Abstract

Analyzing user-generated content from social media can provide insight for organizations into how customers evaluate services and identify potential pitfalls. This paper explores approaches for analyzing content published by users on X with a main goal of finding messages mentioning airport services to use as feedback for improving airport service quality (ASQ). A dataset of over 1 million tweets mentioning international airports was cleaned, explored, and preprocessed for the task of detecting topics related to airport services, using a rule-based technique, and off-the-shelf classification algorithms. The topics are based on the Airport Council International (ACI) ASQ measurement framework. The authors studied the presence of various airport service categories in the messages, paying special attention to accessibility. Thereafter, sentiment analysis was performed to assess customer perception. Finally, five machine-learning algorithms for multi-label classification were experimented with to detect airport services. The study shows that more than 65% of messages in collected data did not mention any airport services. Around 0.6% of messages concern accessibility, primarily focusing on the process of reaching the airport. Three tested models achieve accuracy, comparable to keyword-based labeling. Additionally, a method for evaluating and comparing sentiment across various airport service categories was proposed, based on automatic sentiment analysis scores