人文知识发现的计算机实现——对“汉典古籍”主题建模的实证分析  

To Discover Humanities Knowledge by the Computer: An Empirical Analysis of Topic Modeling the “Handian” Ancient Chinese Classics

作　　者：王小红[1] 科林.艾伦 浦江淮 李颖娜 WANG Xiaohong1,,Colin Allen1, 2,3(1. School of Humanities and Social Sciences, Xi＇an Jiaotong University, Xi＇an, Shaanxi, 710049; 2. Department of History and Philosophy of Science, University of Pittsburgh, PA 15260, USA; 3. Cognitive Science Program, Indiana University, Bloomington, IN 47405, US)

机构地区：[1]西安交通大学人文与社会科学学院,陕西西安710049 [2]匹兹堡大学科学哲学科学史系,美国宾夕法尼亚州匹兹堡15260 [3]印第安纳大学认知科学项目,美国印第安纳州布卢明顿47405

出　　处：《自然辩证法通讯》2018年第4期50-58,共9页Journal of Dialectics of Nature

基　　金：国家社科基金重大项目(项目编号:17ZDA028)子课题的阶段性成果

摘　　要：主题建模方法是当前数字人文研究中一种新的计算方法,本文以其在人文知识研究和学习中的应用——"汉典古籍"主题建模为例,剖析其对人文知识的计算分析所引发的主题客观性呈现、解读语义、新的阅读方式等方法论、认识论问题,最后对其在人文知识领域应用中的认识误区和难点予以揭示。Topic Modeling is a new computing tool in the context of digital humanities. This paper discusses its application for the research and study of humanities knowledge based on topic modeling the ＂Handian＂ ancient classics. We discuss methodological and epistemological questions raised by this application of topic modeling to the computational analysis of knowledge in humanities. Lastly, we reveal some misconceptions and difficulties in the use of topic modeling to humanities research.

关 键 词：数字人文 主题建模 汉典古籍 客观性 语义 发现 

分 类 号：N0[自然科学总论—科学技术哲学]