Decoding customer satisfaction in shared accommodation in historic districts: A dual-configuration analysis based on objective features and subjective perceptions

Shared accommodation in historic districts enriches tourists' local cultural experiences and supports urban regeneration. However, how objective spatial conditions are perceived and associated with satisfaction remains insufficiently explained. This study examines historic-district shared accommodation through a two-stage design integrating large language models and fsQCA. Study 1 uses Tujia reviews, TopicGPT and aspect-based sentiment analysis to identify perceived quality dimensions and classify their Kano roles. Study 2 combines platform-structured and multi-source geospatial data to compare subjective evaluations and objectively observable conditions in high-satisfaction configurations. Results show two core patterns in both subjective and objective configurations: location–service synergy and landscape–service complementarity, alongside partial divergence related to experiential factors such as service interaction. Findings inform differentiated operation and sustainable regeneration of historic districts.