基于专利全引用信息的技术知识扩散特征研究--以石墨烯技术为例

Research Framework of Technical Knowledge Diffusion Characteristics Based on Patent Full Citation Information:The Example of Grapheme

提出了一个建立在专利全引用信息基础上的技术知识扩散特征研究框架,运用专利引用和文本挖掘的方法,通过研究施引专利和被引专利,从技术知识的利用和技术知识的传播两个角度,更全面地探索技术知识扩散的特征。首先提取基本专利和引文专利之间的引用关系,获得技术知识扩散的范围;其次利用文本挖掘的方法抽取专利摘要中的技术关键词,并计算基本专利与引文专利间的文本相似度,衡量出技术知识扩散的程度;然后运用文本聚类和语义分析法进行专利的技术知识扩散特征分析,提炼出6种技术知识扩散特征下分别对应的技术主题。从专利文本的内容着手,将其与引用相结合进行技术知识扩散特征的研究,并对尚未出现前向引用关系的专利进行了技术知识扩散潜力的测度,减少了以往研究的局限性。最后用石墨烯技术作为实证进行了研究,验证了该方法的可行性和有效性,也为石墨烯技术的发展提供了决策支持。

This paper proposes a research framework of technical knowledge diffusion characteristics based on patent full citation information.We explore the technical knowledge diffusion characteristics comprehensively by researching citing patents and cited patents from the perspectives of the use of technical knowledge and the spread of technical knowledge.First of all,we extract the citation relation between basic patents and citation patents to obtain the scope of technical knowledge diffusion.Then,we extract the technology key words in patent abstracts with the method of text mining,and calculate the text similarity between the basic patents and the citation patents to measure the degree of the diffusion of the technical knowledge.Next,with the method of text clustering and semantic analysis,we carry out the patents’technical knowledge diffusion characteristics analysis,and extract six kinds of topics of different technical knowledge diffusion characteristics,respectively.Starting from the patent text content,it is reducing the limitation of previous studies to combine it with citation to explore technical knowledge diffusion characteristics and measure the potential of technical knowledge diffusion of patents without forward citations.Finally,we carry out an empirical study with grapheme technology,verify the feasibility and effectiveness of this method,and provide decision support for the development of grapheme technology.