摘要
古籍是中华民族历史和传统文化的重要载体,古籍研究对传承传统优秀文化、保护民族文化遗产等具有重要意义.选择先秦儒学典籍《论语》、《孟子》和《荀子》为研究对象,基于主题数优化的LDA模型,提取先秦儒学典籍的主要话题;针对动态主题模型在主题演化分析中的不足,采用基于主题语义关联度的Baseline方法,分析主题在不同时期的演化趋势.相对于动态主题模型难以准确抽取主题演化趋势,Baseline方法发现先秦儒学典籍在“政”、“仁”、“礼”、“教育观”和“认识论”等主题演化趋势明显,且与LDA独立分析结果的趋势一致,验证了Baseline方法在古籍主题演化分析的有效性,实现了先秦典籍的主题提取以及深层次的主题演化分析.
Chinese ancient books are important carriers of Chinese history and traditional culture,and the study of ancient books is of great significance for inheriting traditional excellent culture and protecting national cultural heritage.This paper selects the pre-Qin Confucian classics“Analects of Confucius”,“Mencius”and“Xunzi”as research objects,and based on the topic number optimized LDA model,extracts the main topics of the pre-Qin Confucian classics.In view of the shortcomings of dynamic topic models in topic evolution analysis,a Baseline method based on topic semantic correlation is applied to analyze the evolutionary trends of topics in different periods.Compared with dynamic topic models that are difficult to accurately extract the trends of topic evolution,the Baseline method finds that the evolutionary trends of pre-Qin Confucian classics in topics such as“politics”,“benevolence”,“rituals”,“educational views”and“epistemology”are obvious,and the trends are consistent with the independent analysis results of LDA,which verifies the effectiveness of the Baseline method in the analysis of ancient book topic evolution and achieves the extraction of pre-Qin classics topics and in-depth analysis of Confucian classic topic evolution.
作者
陈进东
刘琳琳
孟祥俊
张健
Chen Jindong;Liu Linlin;Meng Xiangjun;Zhang Jian(School of Economics and Management,Beijing Information Science and Technology University,Beijing 100192;Beijing International Science and Technology Cooperation Base of Intelligent Decision and Big Data Application,Beijing 100192)
出处
《系统科学与数学》
CSCD
北大核心
2023年第8期2103-2115,共13页
Journal of Systems Science and Mathematical Sciences
基金
国家重点研发计划项目(2017YFB1400400)
北京信息科技大学师资补充与支持计划(2018-2020)(5029011103)
北京信息科技大学促进高效内涵发展科研水平提高项目(521201090A)资助课题。
关键词
主题模型
知识挖掘
主题演化
儒学典籍
Topic model
knowledge mining
topic evolution
Confucian canonical texts
