This study introduces a novel framework to dissect and understand tourists′ cultural perceptions within river basins. The framework consists of two complementary parts: first, it develops a multi-dimensional system to identify cultural perceptions through textual analysis; second, it uses advanced methods like deep learning and spatial clustering to analyze and compare these perceptions across different cities and regions. The findings from the Yellow River Basin reveal six key dimensions of cultural perception: historical, architectural, folklore, food, religious, and leisure. The basin exhibits three distinct cultural patterns: an upstream polycentric network, a central ‘cultural circle’ around Xi’an, and a city-to-city pattern downstream. Furthermore, the basin is categorized into ten unique cultural perception regions, each highlighting diverse tourist perceptions. This framework not only offers a methodological beacon for future regional tourism studies but also equips managers with strategic insights to enhance the quality and cooperation in river basin tourism development.