MONGKOLCHAI TIANSOODEENON (RAMKHAMHAENG UNIVERSITY)

TEXT MINING APPROACH IN LANGUAGE ANALYSIS TO CLASSIFY AIRLINE PASSENGERS’ RECOMMENDATIONS REVIEWS : DIGITAL LITERACIES/LANGUAGE LEARNING AND TECHNOLOGY

The rapid advancement of artificial intelligence presents exciting opportunities and complexities in language analysis and the management of massive amounts of data. Furthermore, the adoption of an alternative data analysis approach appears to have overshadowed research in language studies. Thus, this study offers additional insight into the application of a text mining approach to analyze passengers’ reviews of service provided by the top ten airlines of the year 2023. It intends to compare the level of accuracy to predict the recommendations of the airlines by using Naïve Bayes and k-NN algorithms. This study programmed a text preprocessing procedure that included word segmentation, stop words, stemming, N-grams, and vectorization (TF-IDF, term frequency, term occurrences, and binary term occurrences). The dataset involved 527 passengers’ reviews of the top-ten best airline of the year 2023 from the customer reviews section of the SKYTRAX website. The findings revealed that using the Naïve Bayes algorithm, TF-IDF, and two N-grams to sort from the text preprocessing procedure gave them the most accurate predictions. In this context, a label of Y signifies "recommended," whereas N denotes "not recommended." This study has the potential to reduce human error in data classification and enable airlines to optimize their passenger handling procedures more effectively.

Nutthapat Kaewrattanapat, Ph.D. (first author), holds the position of assistant professor in digital technology for education at the Faculty of Education at Suan Sunandha Rajabhat University. He is interested in data science, educational data mining (EDM), and design-based research in digital technologies. Mongkolchai Tiansoodeenon, Ph.D. (corresponding author), is an English lecturer at the Department of English and Linguistics, Ramkhamhaeng University. His areas of interest involve teaching methodology, English for specific purposes, corpus linguistics, and lifelong learning.