Great Rebalancing

This dataset contains the cleaned corpus, topic‑modelling outputs, and theory‑mapping materials used in the study on AI and higher education from 2000–2025. It includes bibliographic metadata for 4,280 publications, BERTopic cluster assignments, the RCT–HFT theoretical lexicon, topic–theory similarity scores, and yearly efficiency/humanity indices used for temporal analysis. No full texts are included—only metadata and derived indicators—to support replication and further research on AI governance and human‑centred educational development.

本数据集包含2000年至2025年间人工智能（AI）与高等教育研究中所使用的清洗后语料库、主题建模输出结果以及理论映射材料。该数据集涵盖4280篇文献的书目元数据、BERTopic聚类分配结果、RCT-HFT理论词典、主题-理论相似性得分，以及用于时序分析的年度效率/人文指数。本数据集未包含完整文本，仅提供元数据与衍生指标，旨在支持人工智能治理与以人为中心的教育发展领域的可复现研究及后续探索。