企业关联与风险传导数据集

数据内容：本数据集属于“企业关联与风险传导”研究数据。数据涵盖区域股权市场中企业的公司基本信息、股权关联，以及各省份企业的详细数据。企业股权关联数据开始于2018年，时间范围上主要集中在2023年及之后的数据。空间范围上覆盖全国31个省份，且呈现出在大城市中更为集中的特点。企业专利数据按年份统计，最新更新至2023年。
数据来源：原始数据来源于CSMAR数据库，经过严格的数据清洗和整合处理后生成
分析手段：首先对企业分布、股权关联网络和专利类型分布进行统计分析，揭示企业间关联的空间特征与规模效应。构建基于股权关联的企业网络，使用Python中的NetworkX分析网络的中心性、密度、节点影响力等关键指标。基于VAR（向量自回归）模型和Granger因果检验分析企业间股权关联如何影响风险传导。采用面板数据模型或多层线性模型，研究专利关联的强度与企业创新绩效之间的关系。
对考核指标的支撑性：本数据集能够支持对企业间关联结构及风险传导机制的深入分析。
数据集之间的关联:本数据集和数据集(18)(19)(20)(21)(22)(24)(26)根据本课题任务书《跨链数据可信治理挖掘及数据质量穿透智控》中的关键技术问题与研究目标，旨在针对区域性股权市场数据本体多样、内容丰富、数据维度超高等特点，设计跨链数据治理方案，建立科学有效的区块链数据治理体系，实现对数据来源和安全性的统一标准。
采集方案：数据直接提取自CSMAR（中国证券市场与会计研究数据库），该数据库具有权威性和广泛的学术使用场景，确保了数据质量与可信性。使用CSMAR数据库的导出功能按需求提取相关数据字段，包括企业代码、股权关联比例、专利数量及分类信息等。
时间及地点：2024年8月23日，南京大学
设备情况：客户端操作系统为Windows 10 专业版（64位），内存16GB，处理器为11th Gen Intel(R) Core(TM) i7-11700 @ 2.50GHz；服务器操作系统为CentOS Linux release 7.9.2009（64位），内存32GB，处理器为Intel(R) Xeon(R) Gold 5218 CPU @ 2.30GH。

Data Content: This dataset belongs to the research data of 'Corporate Connections and Risk Conduction'. It covers basic corporate information, equity connections of enterprises in the regional equity market, and detailed enterprise data across all 31 provinces in China. The corporate equity connection data dates back to 2018, with the main focus on data from 2023 onwards. The spatial coverage includes all 31 provinces across the country, and it exhibits a concentration trend in major cities. Corporate patent data is counted annually, with the latest update up to 2023.

Data Source: The original data is sourced from the CSMAR database, and the finalized dataset is generated after strict data cleaning and integration processing.

Analytical Methods: First, statistical analyses are conducted on enterprise distribution, equity connection networks, and patent type distribution to reveal the spatial characteristics and scale effects of inter-firm connections. An enterprise network based on equity connections is then constructed, and NetworkX in Python is used to analyze key network metrics including centrality, density, and node influence. The VAR (Vector Autoregression) model and Granger Causality Test are applied to examine how equity connections between firms affect risk conduction. Panel data models or hierarchical linear models are adopted to study the relationship between the strength of patent connections and corporate innovation performance.

Support for Assessment Indicators: This dataset can support in-depth analyses of inter-firm connection structures and risk conduction mechanisms.

Inter-dataset Association: This dataset, together with datasets (18), (19), (20), (21), (22), (24), and (26), aligns with the key technical issues and research objectives outlined in the project task document *Trusted Governance Mining of Cross-chain Data and Intelligent Control of Data Quality Penetration*. Aiming at the characteristics of diverse ontologies, rich content, and ultra-high data dimensions of regional equity market data, we design a cross-chain data governance scheme, establish a scientific and effective blockchain data governance system, and achieve unified standards for data sources and security.

Data Collection Plan: Data is directly extracted from the CSMAR (China Stock Market & Accounting Research) Database, which is authoritative and widely used in academic scenarios, ensuring data quality and credibility. Relevant data fields including enterprise codes, equity connection ratios, patent quantities and classification information are extracted on demand using the export function of the CSMAR database.

Time and Location: August 23, 2024, Nanjing University

Equipment Configuration: The client operating system is Windows 10 Professional (64-bit), with 16GB of RAM and an 11th Gen Intel(R) Core(TM) i7-11700 @ 2.50GHz processor; the server operating system is CentOS Linux release 7.9.2009 (64-bit), with 32GB of RAM and an Intel(R) Xeon(R) Gold 5218 CPU @ 2.30GHz processor.

搜集汇总

背景与挑战

背景概述

本数据集聚焦于企业关联与风险传导研究，涵盖全国区域股权市场企业的基本信息和股权关联数据，时间范围以2023年及之后为主。数据源自CSMAR数据库并经过处理，可用于分析企业网络结构和风险传导机制。