您当前的位置:首页 > 论文详情

基于BERTopic模型的制造业数字化转型模式识别与演化分析——以山东省上市公司为例

Analysis of Digital Transformation Patterns and Evolution in Manufacturing Based on BERTopic Model: A Case Study of Listed Companies in Shandong Province

摘要: 随着制造业数字化转型的加速推进,企业转型的核心已从“业务求新”转变为“全面重塑”。本文利用BERTopic 主题模型,基于16万字1148个段落的山东省29家制造业上市公司数字化转型文本数据,通过对识别出的54个主题进行层次聚类,并通过与两个大模型交互验证的方式,进一步提炼出16个类团以及7大模块,从而构建出制造业数字化转型模式框架。通过演化分析发现,随着时间推移,实践中制造业数字化转型表现为探索期、起步期、高速发展期三个推进阶段,演化路径呈现出“点-线-面”逐步推进的逻辑。根据研究结论,从制定阶段性转型策略、构建数字化能力体系、推进组织变革创三个方面对制造业企业开展数字化转型提出管理启示。

Abstract: With the accelerated digital transformation of the manufacturing industry, the core of corporate transformation has shifted from "business innovation" to "comprehensive reshaping." This study utilizes the BERTopic topic model to analyze 160,000 words of digital transformation text data from 1,148 paragraphs across 29 manufacturing listed companies in Shandong Province. Through hierarchical clustering of the identified 54 topics and further refinement via interaction with two large language models, we extract 16 clusters and 7 major modules, constructing a comprehensive framework for the digital transformation of the manufacturing industry. Evolutionary analysis reveals that the digital transformation process progresses through three stages: exploration, initiation, and rapid development, following a logical progression from "points" to "lines" to "surfaces." Based on these findings, we provide managerial insights for manufacturing firms in three key areas: formulating phased transformation strategies, building digital capability systems, and advancing organizational change.

版本历史

[V1] 2025-03-17 15:27:32 PSSXiv:202503.01346V1 下载全文
点击下载全文
在线阅读
许可声明
metrics指标
  •  点击量54
  •  下载量18
  • 评论量 0
评论
分享
邀请专家评阅
收藏