Online Social Sentiment Analysis: Discerning Customer Satisfaction via Twitter Reviews on Malaysia’s Courier Service

Abstract

As e-commerce continues to grow, customers frequently encounter issues with courier services such as late deliveries, tracking problems, and unsatisfactory customer support. Although courier services are widely used, there is no dedicated platform that helps users choose the best provider based on social media sentiment. This study investigates customer satisfaction with courier services in Malaysia using Twitter reviews, focusing on three prominent providers: PosLaju, GDex, and DHL. A web-based application was developed using the Flask framework in Python and deployed on the Heroku platform. Sentiment analysis was performed using the TextBlob library, which classifies tweets into positive, negative, or neutral categories. TextBlob was selected for its ease of use and effectiveness in handling basic natural language processing tasks such as polarity detection and phrase extraction. The sentiment results were visualized to reveal trends in user feedback. The analysis showed that PosLaju had the lowest positive sentiment, with only 4.1 percent, and a relatively high negative sentiment of 14.9 percent. DHL received the highest positive sentiment at 39.6 percent and 20.8 percent negative, indicating a stronger approval among users. GDex showed moderate levels of both positive and negative feedback. This study provides valuable insight into customer perceptions of major courier services in Malaysia. By highlighting sentiment trends based on real user experiences from Twitter, the platform can assist future customers in selecting the courier service that best matches their needs.