专利标题与技术特征的关系  

Relationship between patent title and technical features

作　　者：傅晨波[1] 邱君瀚 夏镒楠 俞山青[1] 胡晓娟[2] 闵勇 FU Chenbo;QIU Junhan;XIA Yinan;YU Shanqing;HU Xiaojuan;MIN Yong(Institute of Cyberspace Security,Zhejiang University of Technology,Hangzhou Zhejiang 310012,China;Zhejiang Economy Information Center,Hangzhou Zhejiang 310006,China)

机构地区：[1]浙江工业大学网络空间安全研究院,杭州310012 [2]浙江省经济信息中心,杭州310006

出　　处：《计算机应用》2022年第S02期111-117,共7页journal of Computer Applications

基　　金：浙江省基础公益研究计划项目(LGF20F020016,LGF21G010003)。

摘　　要：为了定量化研究中国专利,采用Z分数的方法研究中国专利标题与技术特征之间的关系。首先,采集网上近35万条电通信领域的专利数据构建专利数据集;其次,通过国际专业分类号表(IPC)中分类号的配对计算每项专利的Z分数量化专利技术特征,并统计专利的标题长度,其中技术特征主要指专利的新颖性和典型性;最后,通过统计分析和格兰杰因果关系检验的方法量化研究两者之间的关系。实验结果表明:专利标题长度与专利技术特征的新颖性和典型性之间存在短期的因果关系,新颖性或典型性越强,专利标题长度就越短,P值均小于0.05;此外,当专利同时具有较好的新颖性和典型性的时候,专利影响力最高。这个发现在一定程度上为发明人提升专利技术水平、提升专利影响力提供了一些可供参考的方法。In order to study Chinese patents quantitatively,the relationship between Chinese patent title and technical features by using Z-score method was studied.First,a patent data set by collecting nearly 350000 patent data in the field of telecommunication was constructed.Secondly,the Z-score of each patent was calculated by the pairing of IPC(International Patent Classification table)classification numbers to quantify patent technical features,and the title length of patent was counted,where the technical characteristics mainly refer to the novelty and conventionality of the patent.Finally,the relationship between the two was quantified by methods of statistical analysis and Granger causality test.Experimental results show a short-term causal relationship between the length of patent title and the novelty and conventionality of patent technical features.The stronger the novelty or conventionality,the shorter the length of the patent title,and the P values are all less than 0.05.In addition,when the patent has good novelty and conventionality at the same time,the patent influence is the highest.To a certain extent,this discovery provides suggestions for inventors to improve the level of patent and enhance patent influence.

关 键 词：专利 技术特征 标题长度 Z分数 因果检验 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术]