Scholarlink

Public transportation systems worldwide are increasingly leveraging user-generated content on online platforms to improve service quality. Bangkok's transit system, characterized by its high utilization, provides a wealth of customer feedback that can guide enhancements. This study employs Aspect-Based Sentiment Analysis (ABSA) to dissect customer reviews of Bangkok's train services, focusing on specific service aspects such as cleanliness, punctuality, and staff behavior. Traditional machine learning models—Logistic Regression, Support Vector Machines, Decision Trees, and Random Forests—are implemented and compared to identify the most effective model for sentiment classification in this context. Utilizing a dataset of 93,948 reviews collected over six years, the Random Forest model emerged as the top performer with an accuracy of 88.90%. The findings highlight the continued relevance of traditional machine learning techniques in sentiment analysis, offering practical solutions for transportation service providers aiming to enhance customer satisfaction.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Aspect-Based Sentiment Analysis of Bangkok's Public Transportation Using Traditional Machine Learning Models

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Anh Nguyen Tran,
  • Vinh Van Thanh Nguyen,
  • Karn Na Sritha,
  • Khanista Namee

摘要

Public transportation systems worldwide are increasingly leveraging user-generated content on online platforms to improve service quality. Bangkok's transit system, characterized by its high utilization, provides a wealth of customer feedback that can guide enhancements. This study employs Aspect-Based Sentiment Analysis (ABSA) to dissect customer reviews of Bangkok's train services, focusing on specific service aspects such as cleanliness, punctuality, and staff behavior. Traditional machine learning models—Logistic Regression, Support Vector Machines, Decision Trees, and Random Forests—are implemented and compared to identify the most effective model for sentiment classification in this context. Utilizing a dataset of 93,948 reviews collected over six years, the Random Forest model emerged as the top performer with an accuracy of 88.90%. The findings highlight the continued relevance of traditional machine learning techniques in sentiment analysis, offering practical solutions for transportation service providers aiming to enhance customer satisfaction.