Text Mining: Sentiment Analysis of Reviews on TripAdvisor for Vietnam’s Michelin-Starred and Selected Restaurants in the 2023 Michelin Guide

• Published in: Journal of Multimedia Information System Vol. 11; no. 2; pp. 131 - 148
• Main Authors: Nguyen, Ha Linh, Dao, Thi Hai
• Format: Journal Article
• Language: English
• Published: 한국멀티미디어학회 30.06.2024
• Subjects: 컴퓨터학
• ISSN: 2383-7632; 2383-7632
• DOI: 10.33851/JMIS.2024.11.2.131

The proliferation of the Internet has led to a surge in online reviews, making social media platforms a crucial resource for consumers before making purchases. This study examines how the overall sentiment towards Vietnam’s restaurants changes after they receive the recognition in the 2023 Michelin Guide, as well as which dimensions (food, service, ambiance and price) experience the greatest or least alterations after the award and which dimensions most influence the overall sentiment. The research involves a sentiment analysis of online reviews of Vietnam’s restaurants honored by the 2023 Michelin Guide on TripAdvisor from 1 January to 31 December 2023. A total of 1,292 English and Vietnamese reviews from 4 Michelin-starred restaurants and 70 Michelin-selected restaurants located in Hanoi and Ho Chi Minh City were extracted using a web crawler named Apify. Data were then processed and analyzed by using data collection methods, text mining methods, and the Azure sentiment analysis tool in Excel. The research findings reveal that the overall sentiment declined after the Michelin recognition. Among the four dimensions, food was the most affected criterion, followed by service, ambiance, and price. Notably, the sentiment around ambiance was the highest and increased after the award. Overall sentiment towards Michelin-selected restaurants is higher than that of Michelin-starred restaurants. These insights provide valuable guidance for Michelin restaurant owners facing challenges in maintaining quality, meeting high customer expectations, managing their online reputation, and for diners intending to eat out to choose a suitable destination. This first research about Vietnamese Michelin restaurants using text mining and sentiment analysis methods identifies specific areas for improvement in food, service, ambiance, and price, and highlights the different impacts on Michelin-starred versus Michelin selected establishments, offering targeted recommendations to enhance services, attract more positive reviews, and improve overall performance and customer satisfaction. KCI Citation Count: 0