【摘要】 如何从大规模文本中抽取和泛化事件已成为当前古籍事件研究的一个关键问题。针对古籍文本和古代汉语的特点，本文构建了一种半监督事件聚类模型USKm,该模型利用USIF表征古文历史事件，基于约束距离集成，将邻近区域点纳入数据点类簇的二次决策过程对事件进行聚类从而实现主题识别。以《后汉书》为研究对象，笔者对比了USKm与传统聚类模型的应用效果，发现USKm性能更优。笔者可视化东汉政权存续期间时间分布，绘制历史事件人物关系图谱，并解析背后的历史现象探讨东汉政权的发展规律。USKm模型通过半监督训练，提高了事件特征的识别准确性和聚类效果，同时本文对聚类结果数据加工整理与可视化，从数字人文视阈为人文研究者提供新的研究思路和角度。更多还原

【Abstract】 Extracting and generalizing events from large-scale texts has emerged as a crucial issue in the contemporary studies of ancient literature. Based on the distinctive characteristics of ancient texts and ancient Chinese language, this study introduces a semi-supervised event clustering model named USKm, which employs USIF to represent historical events found in ancient texts. By incorporating neighboring data points into the secondary decision-making process, the model effectively achieves the goal of event grouping and topic recognition. Using the Later Han Shu as the primary subject of investigation, the author compares the performance of USKm with traditional clustering models, revealing its superior effectiveness. The study further visualizes the temporal distribution of the Eastern Han Dynasty, constructs a graphical depiction of the interplay between historical figures and events, and delves into the underlying historical phenomena to unveil the developmental trends of the period. The USKm model enhances the precision of event feature recognition and clustering efficacy through semi-supervised training, offering novel research insights and perspectives for scholars in digital humanities by processing and visualizing the clustered data results.更多还原