基于指数随机图模型的专利技术扩散机制实证研究

网络嵌入视角下专利技术扩散路径间存在相关性，但这种网络内生性影响并未在已有研究中得到足够重视。将扩散因素归纳为：专利价值、专利保护、多维同配、结构嵌入，从扩散网络内生、外生因素两个方面构建分析模型，并以石墨烯领域为例，采用指数随机图模型，开展实证研究。实证结果说明：专利价值、技术、地理、组织层面同配性、闭合三角结构嵌入均正向促进专利技术扩散，而专利保护强度反向抑制技术扩散。值得注意的是，专利技术扩散存在强烈的结构嵌入倾向，融合网络内、外生因素的模型能够更好地捕捉专利技术扩散机制。

Empirical Research of Patent Technology Diffusion Mechanism Based on ERGM

Dependency relation exists between diffusing path of patent technology from view of network embeddedness, while this kind of endogenous effects from network has not been fully investigated.Factors influencing patent technology diffusion were classified into patent value, patent protection, multidimensional assortativity, and structure embeddedness.And analyzing model was formulated by dividing the variables with respect to influencing factors into two parts involving endogeneous and exogeneous variable.Moreover, graphene patents was chosen for empirical research in order to validate theoretical hypothesis by using method of ERGM.Therefore, empirical conclusions were drawn that patent technology diffusion is positively correlated with patent value, multidimensional assortativity with respect to technology, geography, society, embeddedness of closed triangle structure, while negatively correlated with intensity of patent protection.It is worthwhile to pay attention to the intense impact of structure embeddedness, and the model incorporating all factors involving endogenous and exogenous issues is capable of stronger capturing capability for the mechanism of patent technology diffusion than another with only a part of all issues.