Exploring Customer Experience and Sentiment Analysis among By.U Users

Internet users' experience has significantly increased because technology is growing in the digital age, especially social media. With easy access to Twitter and Instagram, this study aims to determine the customer experience of by.U users to realize customer-centricity. This study uses sentiment analysis, multi-class classification, and topic modeling using a BERT deep learning algorithm. In comparison, topic modeling is used to see the insight of context and the topic as a whole. The results showed 89% accuracy on sentiment analysis and 85.4% accuracy on multi-class classification. This study found that by.U got the highest negative sentiment on the service quality dimension. The analysis results in this study can provide benefits for companies to find out the dimensions of customer experience that need to be improved by the company.