Online User-Generated Content (UGC) offers valuable insights into the urban vitality of specific districts, with travelogues revealing tourists’ overall preferences and informing analyses of urban functionality. However, translating large-scale textual data into geographically anchored, preference-based information remains challenging due to the abstract and often weak spatial link in narrative descriptions. This study proposes a UGC-driven Agent-Based Modeling (ABM) workflow that integrates geocoded location data and visitation preference analysis into agents’ decision-making processes, enabling a bottom-up simulation of tourist behaviors derived from textual sources. Using the Pingjiang Historical and Cultural District as a case study, 16,053 travelogue entries (2009–2023) were analyzed, encoded, and simulated across four control groups. Results show that the UGC-driven ABM effectively reproduces movement patterns and vitality distributions, with traffic-related outputs aligning closely with Space Syntax analyses and POI-based heatmaps reflecting tourist preferences. The findings demonstrate that this approach provides a practical and scalable method for extracting and spatializing behavioral insights from textual data, offering applications in urban planning and tourism management.
Zhang et al. (2026) studied this question.