基于典籍的春秋社会时间序列演变分析方法初探

Social Changes in Period Chunqiu of China:From the Perspective of Time-Series Analysis of Chinese Classical Book

春秋时代作为中国古代历史发展的重要转型时期,经济、政治、文化等各领域都发生了急剧的变革。研究这一时期社会变迁,对于阐释中华文化的历史渊源、发展脉络、基本走向,建构中国特色社会主义传统文化观皆有非常重要的意义和价值。文章以《左传》为语料来源,以文本挖掘为手段分析春秋社会演变规律,借助社会变迁相关理论对春秋时代进行结构、表现及动力等不同维度的描述,进而从文本分析的角度构建对应的量化指标。通过融合词频分析、聚类分析、时间序列分析、社区结构挖掘等多种文本挖掘技术,实现各项量化指标的计算。实验结果表明,研究设计的文本计算方法较好地描述了春秋社会结构演变、演变动力及演变表现,与人文学者研究结果基本一致。对于人文计算的开展具有一定理论价值与实践意义,但在模型构建、特征挖掘的方法以及结果评价方面仍有待进一步提升。

Period Chunqiu(B.C.770-B.C.476)is one of the most important social transition and revaluation periods in the development of Chinese ancient history.There were drastic changes in the domain of economics,politics,cultures etc.in that period.To research the social changes in this period has great significance and values not only to explain the origin of the history,the historical development and the development of tendency of Chinese nation,but also to illuminate the formulate of traditional cultural view of socialism with Chinese characteristics.The object of the paper is to analysis the social changes regulation in Chunqiu Period by the method of text mining.In order to that,firstly,we build a theoretical model to describe the development of social changes in different dimensions of structure,expression and dynamics by using the relevant theory of social change.And then some indicators were refined to figure the dimensions we designed in the theoretical model in the framework of texting mining and analysis.Later,all of the indicators were figure out by using hybrid techniques of words frequency analysis,cluttering analysis,time series analysis and network community mining.The text mining results reveal the social changes in Period Chunqiu in the changes of social structure,dynamics and expression from the perspective of humanity computation.The paper will inspire the research of humanity computation by using text mining and analysis technique.However,some progress could be achieved by building a more comprehensive model,and improving the precision of text feature mining as well as the evaluation in texting mining results.