Key points are not available for this paper at this time.
The airline industry relies heavily on employee satisfaction and customer perception to drive financial success, yet the interconnected impact of these factors on performance remains underexplored. This study introduces a generative AI (GenAI)-assisted text analytics and Data Envelopment Analysis (DEA) framework to evaluate airline efficiency by linking employee satisfaction, customer sentiment, and financial outcomes. Leveraging recent online reviews as a source of unfiltered feedback, the proposed three-stage methodology includes: (i) extracting and categorising key service quality and operational indicators using BERTopic, a state-of-the-art topic modelling technique, augmented with GenAI for improved topic labelling and classification; (ii) analysing sentiment polarities with GenAI-assisted sentiment analysers; and (iii) linking these insights to financial performance using fuzzy network DEA (FNDEA). A case study involving eight US-based airlines and 25,000 customer reviews alongside 24,000 employee reviews is used to evaluate the proposed framework. Experimental results reveal that GenAI-assisted techniques significantly outperform traditional topic classification and sentiment scoring methods. Findings also highlight efficiency disparities among airlines, with those achieving higher employee satisfaction and positive customer sentiment demonstrating superior financial outcomes. This framework offers actionable insights for identifying strengths and addressing weaknesses, enabling airlines to strategically improve service quality, operational efficiency, and competitiveness.
Raad et al. (Fri,) studied this question.