Text mining-based analysis of ancient landscape and tourism behavior at Hangzhou’s West Lake

Poetic texts are crucial for uncovering heritage sites’ cultural values. This research uses natural language processing (NLP), Python-based co-occurrence semantic networks, and kernel density analysis to study 2065 ancient poems on Hangzhou’s West Lake, exploring landscape imagery and sightseeing behaviors. It identifies four typical landscape images with specific elements, structural characteristics, and significant spatial differentiation, influenced by natural foundations, socio-economic contexts, and ideologies, which shape diverse tourism motivations and tourism behaviors. Based on these insights, it proposes conservation and utilization strategies at individual sites, tourist routes, and regional levels to enhance preservation and sustainable use, aiding in restoring original imagery, preserving authenticity, and providing valuable insights for heritage management amid modern tourism demands.