A large language model-based tourist satisfaction index

We propose a novel framework for compiling a tourist satisfaction index using online reviews and large language models. Building on a theoretical foundation linking tourism service quality with tourists' cognitive–affective responses and satisfaction, semantic analysis of online reviews is conducted to assess tourist satisfaction. Unlike traditional questionnaire-based methods, this framework offers a dynamic, data-driven means of evaluating tourist satisfaction. It facilitates timely assessment of tourist satisfaction with tourism products and services at the business, sectoral, and destination levels and enables segment analyses of trip and spatial characteristics. Thanks to these features, the novel framework holds considerable potential for effective, targeted, and scalable destination management.