基于可持续发展报告披露信息挖掘的企业财务绩效研究：供应链视角

The release of Global Reporting Initiative Sustainability Reports (GRI SR) and the development of new text-mining techniques such as natural language processing (NLP) and machine learning (ML) are providing development opportunities for the management of corporate financial performance. However, there are great challenges in matching the need for global, precise and timely financial-performance management with the typical characteristics of GRI SR disclosure (e.g., multiple-source information, discrepant structures, intense data noise, sparse effective data, and dynamic and fragmented). This study uses this context to explore the interplay between corporate social responsibility report announcements and financial-performance management through text-mining techniques from the perspective of supply-chain management. Specifically, the study has the following four objectives: (1) construct a multiple-attribute analytical framework of corporate GRI SR and its linguistic database; (2) extract the “supply-chain characteristic word-opinion duad” attributes of GRI SR; (3) identify the association between corporate financial performance and the supply-chain characteristic word-opinion duad attributes of GRI SR; (4) develop the original system of new types of corporate-information disclosure from the perspective of supply-chain management. This study attempts to address the key issues in the deep integration between GRI SR information disclosure and financial performance from the perspective of supply-chain management to comprehensively promote the financial-performance-management capability of firms and to achieve the initiatives of multidimension, multilayer and sophisticated decision making. This study will provide significant theoretical and practical contributions to building new types of data-driven information-disclosure systems, and to improving the current situation of Chinese firms in fulfilling their social responsibilities.

企业可持续发展报告(GRI SR)与新一代文本挖掘技术的发展为企业财务绩效管理革新创造了契机。然而，GRI SR多源异构、高噪稀疏、动态碎片的典型特征与财务绩效管理全局性、精准性、及时性的新需求也带来了挑战。本课题基于供应链管理视角，应用自然语言处理、机器学习等文本挖掘技术对社会责任报告披露与财务绩效管理展开研究，具体内容如下：第一，构建企业GRI SR多属性分析及其语料库；第二，抽取GRI SR“供应链特征词-观点对”特征；第三，识别GRI SR“供应链特征词-观点对”特征与企业财务绩效关系；第四，开发供应链管理下企业新型信息披露原型系统。本课题力图从供应链管理视角解决GRI SR信息披露与财务绩效深度融合的关键问题，全面提升企业财务绩效管理能力，实现多维度、多层次、精细化决策主张。研究成果能够协助构建数据驱动的新型信息披露系统，提升我国企业履行社会责任水平,助力可持续发展。

企业可持续发展报告(GRI SR)与新一代文本挖掘技术的发展为企业财务绩效管理革新创造了契机。然而，GRI SR多源异构、高噪稀疏、动态碎片的典型特征与财务绩效管理全局性、精准性、及时性的新需求也带来了挑战。本课题基于供应链管理视角，应用自然语言处理、机器学习等文本挖掘技术对社会责任报告披露与财务绩效管理展开研究，具体内容如下：第一，构建企业GRI SR多属性分析及其语料库；第二，抽取GRI SR“供应链特征词-观点对”特征；第三，识别GRI SR“供应链特征词-观点对”特征与企业财务绩效关系；第四，开发供应链管理下企业新型信息披露原型系统。本课题力图从供应链管理视角解决GRI SR信息披露与财务绩效深度融合的关键问题，全面提升企业财务绩效管理能力，实现多维度、多层次、精细化决策主张。研究成果能够协助构建数据驱动的新型信息披露系统，提升我国企业履行社会责任水平，助力可持续发展。