Understanding Visitor Sentiment of Batu Cave Destination through TripAdvisor and Vlogger Content Reviews

Abstract

This study utilizes the CRISP-DM framework to conduct a comprehensive sentiment analysis of visitor reviews for Batu Cave, leveraging advanced tools such as VADER, TextBlob, and the SVM model. The analysis of 1201 TripAdvisor reviews reveal critical visitor perceptions, highlighting both positive aspects, such as the site's beauty and cultural significance, and areas needing improvement, including accessibility and visitor conduct. The SVM model demonstrates high performance with an accuracy of 94.25% and AUC scores of 0.966 (optimistic), 0.962 (standard), and 0.958 (pessimistic). Furthermore, toxicity scores from the Perspective API range from 0.05055 to 0.89882, identifying areas for enhancing visitor interactions. These findings underscore the importance of using data-driven approaches to improve destination management and visitor satisfaction. The study provides valuable insights for policymakers, guiding strategic planning and sustainable development of tourist destinations. Consequently, the research offers a robust foundation for informed decision-making in the tourism sector, aiming to enhance the overall visitor experience at Batu Cave.

Keywords

• batu cave
• crisp-dm
• malaysia
• sentiment
• tripadvisor