Topic Modeling and Sentiment Analysis using Online Reviews for Bangladesh Airlines

Abstract

Air travel is one of the most used ways of transit in our daily lives. So it's no wonder that more and more people are sharing their experience with airlines and airports using web-based online surveys. This study aims to do topic modeling and sentiment analysis on Skytrax (airlinequality.com) and Tripadvisor (tripadvisor.com) postings where there is a lot of interest and engagement from people who have used it or want to use it for airlines. The goal of individuals gathering at Skytrax and Tripadvisor is to make better decisions based on the actual experiences of other customers who have flown with airlines. We gathered online reviews submitted by consumers who have flown with Bangladesh airlines in the past. The data was collected from internet reviews from 4 Bangladesh airlines, totaling 1095 reviews. With the obtained data, topic modeling and sentiment analysis were utilized to determine key phrases in the online reviews. Through frequency analysis, the topic modeling revealed that ‘airline,’ ‘service,’ ‘crew,’ and ‘food’ were essential concerns in the trip. Furthermore, the results indicated that the primary issue, which may impact customer discontent, was delay, but ‘food’ can make customers satisfied through sentiment analysis, as the result indicates the ‘Food and beverage’ topic with no. of highest word count ‘good’, ‘biman’ and ‘service’ keyword in the topic modeling.