网络分析与文本挖掘视角下的宋代士人交游研究OA

[Purpose/significance]The Song dynasty ranks among the most flourishing periods for literati culture in an-cient China.This paper aims to observe the interaction between the literati's friendship network,social culture,and changes in the political system in the Song dynasty from the perspective of digital humanities.[Method/process]This paper uses the China Biographical Database(CBDB)to create a high-quality dataset of cleansed historical elite group in the Song dynasty for social network analysis and network topology statistics.By connecting CBDB with The Com-plete Collection of Song Poems,we build a dataset and a corpus of Song literati's friendship poem.The paper then uses word vector algorithms to establish a semantic network of feature words in the poems,and identifies the topics of the friendship poems with community detection algorithm.[Result/conclusion]The results show that the literati group formed various social activities through all kinds of social relations and opportunities,and the network exhibits the so-called small world property of current social media networks.9 top communities identified from the corpus reflect the fact that social activities were highly encouraged among the literati during the Song dynasty.