A community detection method towards analysis of Xi Feng parties in the northern song dynasty

Factional conflict has played an important role in historical research, espcially studying the features of political scholars. Previous studies mainly analyze factional conflict based on hand-work, which makes it difficult to consider minor political scholars. In this paper, we use the Chinese Historical Biography Database to build a graph based on the relations among scholars in the Xi Feng factional conflict, and then use social network algorithms to model and analyze their impact. We use community detection algorithms to unsupervisedly extract the clusters of the two major parties. By analyzing the data obtained from the algorithm, we confirms several conclusions about the political influence of Cai Jing, Wang Anshi, and the new and old parties in the study of North Song history.