Large Language Model-based Topic-Level Sentiment Analysis for E-Grocery Consumer Reviews

Customer sentiment analysis plays a pivotal role in the digital economy by offering comprehensive insights that inform strategic business decisions, optimize digital marketing initiatives, and improve overall customer satisfaction. We propose a large language model-based topic-level sentiment analysis framework: a BERT model is used to obtain vector representations of documents, and then clustering algorithms are automatically applied to group documents into topics. Once the topics are formed, a GPT model is used to perform sentiment classification on the content related to each topic. The simulations show the effectiveness of this approach, where choosing the proper clustering technique can produce more semantically coherent topics. From a practical perspective on the Indonesian e-grocery customer reviews, the framework identifies unique customer concerns that critical for e-grocery customer satisfaction. Furthermore, topic-level sentiment polarization uncovers that 31.7% of all negative sentiment concentrates around the shopping experience topic despite an overall positive sentiment trend.