NPCN:基于向心引用网络的专利被引频次标准化方法研究

NPCN: A New Method of Patent Citations Normalization Based on Ego Patent Citation Networks

[目的/意义]专利的被引频次是衡量专利影响力的重要指标。由于专利的引用潜力因学科和发表年份的不同而呈现出巨大差异,因此专利的被引频次需要标准化之后才可以实现跨学科、跨年份比较。[方法/过程]基于专利向心引用网络构建一个新的专利被引频次的标准化指标——NPCN。为了验证该指标的有效性,从Dimension数据库中获取2005-2010年3D打印领域已获授权的专利,将其按照FoR分类进行学科划分,分析不同学科、不同年份的专利在被引频次和NPCN分布情况,并采用比均值法、比参考文献法、Z score、NPCN对比它们与专利被引频次的相关关系。[结果/结论]3D打印领域的专利在22个FoR学科分类中均有分布,但不同学科、不同年份的专利在被引频次上差距较大,经过NPCN标准化处理之后,不同专利之间差距变小,呈现出明显的同分布趋势。在相关性方面,NPCN相对其他标准化指标来说与被引频次的相关程度低。

[Purpose/Significance] The number of citations received by patents is an important indicator to measure the influence of patents. Since the citation potential of patents varies greatly by disciplines and publication years, the number of citations received by patents needs to be normalized before crossdisciplinary comparison and cross-year comparison. [Method/Process] Based on ego patent citation network,we constructed a new method, NPCN, to normalize patent citations. Besides, we took patents granted from 2005 to 2010 of 3D printing indexed in Dimensions to verify the effective of NPCN. Specially, we divided these patents into different disciplines with Fields of Research(FoR) and compared the distribution of patent citations and NPCN of patents in different disciplines and publication years. Besides, we also selected the mean method、reference patents method、Z-score method and NPCN methods and compared them with each other in correlation with patent citations. [Results/Conclusions] Patents of 3D printing are categorized into 22 FoR. And it is different between patents in the number of citations. After normalized, the difference between patents in different disciplines and publication years is smaller and a normalized citation distribution is shown. In correlation with patent citations, NPCN is less correlated with patent citations than other normalized methods.