基于文本挖掘方法的环境信息披露指标体系改进——以川渝经济区上市企业为例  

Improvement of Environmental Disclosure Index Based on Text Mining——The Empirical Analysis of Sichuan and Chongqing Listed Companies

作　　者：张秀敏[1] 杨连星[2] 吴漪[1] 曹柳龙 徐一方[1] 

机构地区：[1]华东师范大学商学院,上海200062 [2]中国人民大学经济学院,北京100872

出　　处：《软科学》2014年第7期65-71,共7页Soft Science

基　　金：国家社会科学基金青年项目(10CGL049);中国博士后科学基金特别资助项目(201104649)

摘　　要：对国内外文献中环境信息披露指标(EDI)的构建方式进行系统的梳理,阐释内容分析法在构建质与量相结合指标体系中存在的局限性,提出采用文本挖掘方法确定指标体系中三级指标的内容及权重。在介绍文本挖掘方法生成五种EDI指标体系基础上,结合国内外文献中典型的影响因素,分析EDI与财务绩效之间的关系,通过面板回归得出与既有文献较为一致的结论。Based on the systematic review of the construction method of Environmental Disclosure Index （EDI） from the present domestic and overseas literature, this paper finds the limits of content analysis method in constructing the EDI system based on quantitative and qualitative information, and proposes a complementary method of Text Mining to specify the thirdlevel indicators and their responding weights. After introducing five types of EDls using this method, it selects some typical determinants from the present literature, and start a panel data analysis to analyze the relationship between EDI and financial performance.

关 键 词：环境信息披露 指标 文本挖掘 财务绩效 

分 类 号：F127[经济管理—世界经济] F275