Exploring knowledge structure of "dual carbon" policies : combining computational text mining and knowledge graph

To support energy transition and climate action, China has announced its "dual carbon" goals and introduced a series of related policies. However, these policies remain disorganised and scattered across various platforms and formats, posing significant challenges for accessibility and effective utilisation. To address this, we propose an integrated approach that combines computational text mining with knowledge graph construction to uncover the underlying knowledge structure of the policy systems. Specifically, our method considers both the external attributes and internal conceptual content of policies. We develop an adaptive knowledge extraction framework that incorporates multiple text mining techniques. By examining the network structure of the resulting policy-oriented knowledge graph, we identify key patterns and problems in the current policy system spanning dimensions such as citation networks, institutional roles, legal force, thematic focus, objectives, and implementation measures. Results indicate that the top-level documents play both a leading and bridging role, and that China is complementing administrative directives with increasing use of market and financial instruments. The analysis also highlights areas for improvement, including strengthening inter-provincial coordination, enacting specialised legislation, and refining the specificity and standards of policy implementation. The insights offer valuable strategic guidance for Chinese policymakers while offering transferable implications for other nations pursuing climate and energy strategy goals.