An evaluation index system for China’s “dual carbon” policy was constructed through text mining, the PMC index model was selected combining with self-coding (AE) technology for indicator weight quantification and multi parameter fusion, and 9 primary indicators and 39 secondary indicators were used to achieve quantitative evaluation of 10 representative policies in China’s “1+N” dual carbon policy system. The results indicate that the “dual carbon” policy can be divided into three levels, which exhibit a top-down characteristic. The main reasons for the difference in scores include policy nature, policy timeliness, incentive protection, and policy objects. Based on these findings, feasible suggestions for adjusting and optimizing the “dual carbon” policy have been proposed.