Systematic Literature Review on Sentiment Analysis in Airline Industry

Khan Md. Hasib, Usman Naseem, Ashfia Jannat Keya, Shithi Maitra, Kamruzzaman Mithu, Md. Golam Rabiul Alam

Abstract: The significance of sentiment analysis lies in its ability to identify key areas of customer satisfaction and dissatisfaction, revealing patterns that can guide enhancements in airline services and strategic decision-making. Additionally, the identification of topics within airline passengers’ opinions across various flight classes is crucial, achieved through topic modeling. By analyzing these opinions, we can discern passengers’ priorities and expectations, which are essential for delivering superior service. This survey converges the sparse implications and provides a dense, comprehensive understanding of the state of the art through a Systematic Literature Review (SLR). Our SLR breaks down the research process into three distinct phases: Planning (defining research questions and protocol), Conducting (identifying and assessing relevant studies), and Document Review (synthesizing findings). We analyze 60 papers on sentiment analysis and topic modeling to examine the application and effectiveness of machine learning and deep learning methods in the airline industry. Our study highlights critical tasks in airline sentiment analysis, including multilingual sentiment analysis, and provides an overview of popular datasets, their fundamental properties, and the machine learning and deep learning models applied to them. The existing literature is critically examined to uncover both strengths and weaknesses, yielding additional insights into the prevailing research paradigms.

External IDs:dblp:journals/sncs/HasibNKMMA25