Temporal evolution of Chinese peasant uprising triggers via deep causal graphical models

Since ancient times, Chinese peasant uprisings have played a pivotal role in driving social structural transformation and regime changes. From the Chen Sheng and Wu Guang uprising at the end of the Qin dynasty to the Taiping Rebellion in the late Qing, these movements demonstrate distinct periodic and patterned characteristics. However, traditional historiographical research often remains confined to macro-level narratives, lacking systematic modeling and causal inference of triggering mechanisms, particularly underperforming in cross-period and cross-regional explanatory power. With the rise of digital humanities and advancements in artificial intelligence, deep causal graph models offer new methodological tools for historical inquiry. This study constructs a multi-level feature system encompassing climate, economic, political, and social dimensions, and applies a deep causal graph neural network architecture to systematically model and analyze the triggering mechanisms of peasant uprisings across the 2,414-year span from the Pre-Qin to the Qing dynasty. By integrating graph attention networks, bidirectional LSTM temporal modeling, and enhanced PC algorithms, the study successfully identifies core causal paths and critical nodes of peasant uprisings, validates the significant periodic patterns of 60, 120, and 180 years, and reveals the evolving importance of various structural dimensions across different historical periods.