情报科学 ›› 2025, Vol. 43 ›› Issue (8): 10-19.

• 专论 • 上一篇    下一篇

基于LDA-PMC模型的我国数据要素政策文本量化评价研究

  

  • 出版日期:2025-08-05 发布日期:2025-12-12

  • Online:2025-08-05 Published:2025-12-12

摘要: 【目的/意义】本研究基于LDA-PMC模型,量化分析我国数据要素政策的实施效果,并提出优化建议。随 着数据作为新型生产要素的崛起,研究旨在评估现行政策对数据要素市场的推动作用及其问题。【方法/过程】采用 LDA模型提取政策文件中的潜在主题结构,通过词频分析揭示政策的核心议题。随后,结合PMC指数模型对所选 取的代表性政策的整体效果进行量化评估,分析各政策在内容、执行力及覆盖面等方面的表现。【结果/结论】研究 发现,部分发达地区如北京市、上海市和广东省的数据要素政策在推动市场化配置方面表现突出,特别在数据确 权、流通和安全保障上取得了显著进展。而新疆等欠发达地区政策实施效果较弱,需加强政策执行与市场基础设 施建设。【创新/局限】本研究创新性地结合LDA和PMC模型,填补了主题分析与政策效果量化之间的空白。然而, 样本时间跨度较短,未来研究可扩大样本并引入更多评价维度,以提高分析的全面性与准确性。

Abstract: 【Purpose/significance】This study uses the LDA-PMC model to quantitatively analyze the implementation effectiveness of China's data element policy and to propose optimization suggestions. With the rise of data as a new production factor, the study aims to assess the current policies' role in promoting the data element market and their associated issues.【Method/process】Firstly, the LDA model is used to extract the latent topic structures from the policy texts. Core issues of the policies are revealed through word fre⁃ quency analysis. Then, combined with the PMC index model, the overall effect of selected representative policies is quantitatively evaluated, and their performance in aspects like content, implementation and coverage is analyzed.【Result/conclusion】The study finds that policies in developed regions, such as Beijing, Shanghai, and Guangdong, have shown outstanding performance in promoting market-oriented data allocation, especially in areas like data ownership, circulation, and security. However, policies in underdevel⁃ oped regions, such as Xinjiang, show weaker implementation effects and require improvements in policy enforcement and market infra⁃ structure.【Innovation/limitation】This study innovatively combines the LDA and PMC models, filling the gap between thematic analy⁃ sis and the quantitative assessment of policy effectiveness. However, the sample period is relatively short, and future research could ex⁃ pand the sample size and introduce more evaluation dimensions to enhance the comprehensiveness and accuracy of the analysis.