An LLM-Driven Computer Grounded Theory Study on Sense of Belonging to the City in the Chinese Context

In the context of China's fast urbanisation, “sense of belonging to the city” has been considered for years. From the existing studies, many cities use “community” at the city level as a sense of belonging or use the concept of place attachment, while urban identity is sometimes the character of the city itself. Without doing so, there is no concept to define the emotion that people feel towards this complex urban entity composed of location, community, and culture. While there is a specific concept of “sense of belonging to the city” in the Chinese literature, most of these studies focus on migrant workers or non-native citizens. Therefore, this paper developed and applied a computer-assisted grounded theory process to study online discussion data. Driven by LLMs that simulate the three-level coding, theoretical saturation, and re-analysis of manual grounded theory, the process remains interpretable and auditable despite the large research scale. It was implemented through a Python workflow. Finally, the workflow encoded over 23,000 Zhihu comments, and the study ultimately discussed the dimensions and formation mechanism of the concept of "sense of belonging to the city". Methodologically, this work provides a reproducible template through standardised LLM assistance to apply grounded theory at large scale.