Abstract: In the era of digital economy, data has become a key factor of production. An in-depth discussion on the constituent elements and theoretical system of data factor circulation policy can provide theoretical support for the market-oriented practice path of data factor in China at the policy level. This study combines proceduralised grounded theory and large language model to propose an automated policy text coding method with both theoretical rigor and technical advancement. By designing the coding architecture of "text slicing-analog coding-iterative integration-manual inspection-text extraction" and combining the prompt techniques such as result self-confirmation, role prompting, and thought chain, the problem of large model illusion is effectively alleviated, and the analysis efficiency is significantly improved while the coding quality is guaranteed. This method is effectively applied to data factor circulation policy, and identifies six main categories and their correlation, such as data property rights and security governance, infrastructure and technical support, data element market and its ecological construction. Then, it reveals the shortcomings of the current market development from the perspective of supply and demand, and puts forward policy suggestions on coordinating data right confirmation and data opening, improving the depth and breadth of data application, and promoting the construction of data factor market in a hierarchical and subregional manner.

Key words: Data factor circulation policy, Artificial intelligence generated content, Large language model, Grounded theory, Qualitative research