未确知聚类在专利质量评价中的应用

张妮妮 孙胜娟 张永健

摘  要： 随着人们对知识产权的重视，作为其重要表征的专利的数量呈现爆发式增长，然而专利的质量却没有随之增长。大量的低质量专利不但作用有限，反而会造成社会资源浪费和遏制创新。对于专利质量的评价，目前还没有统一的标准。文中首先对国内外的专利质量指标进行分析，选取出对专利质量影响较大的指标，构建专利质量评价指标模型。同时，以钢铁行业相关专利为目标数据集，分别采用未确知聚类和模糊均值聚类算法对目标专利质量进行分析评价。最终，将目标专利数据聚类出不同的级别，得出高质量专利。在聚类过程中，发现未确知聚类算法在效率和准确率上都有良好的表现。

关键词： 专利质量评价; 未确知聚类; 专利数据分析; 评价模型构建; 数据集聚类; 对比实验

中图分类号： TN911?34; TP311.52                文献标识码： A                     文章编号： 1004?373X（2020）08?0143?04

Application of unascertained clustering in patent quality evaluation

ZHANG Nini， SUN Shengjuan， ZHANG Yongjian

（School of Information and Electrical Engineering， Hebei University of Engineering， Handan 056001， China）

Abstract： With people′s attention to intellectual property， the quantity of patents， as an important symbol of intellectual property， has increased explosively， but the quality of patents has not increased with it. A large number of low?quality patents not only play a limited role， but also lead to the waste of social resources and inhibit innovation. Nowadays， there is no unified standard for the evaluation of patent quality. The domestic and foreign patent quality indicators are analyzed， from which the indicators that have a greater impact on the quality of patents are selected to build the patent quality evaluation index model. With the related patents in steel industry as the target data set， the analysis and evaluation of the target patent quality are performed by means of the unascertained clustering algorithm and fuzzy mean clustering algorithm， respectively. The target patent data is clustered into different levels to obtain high?quality patents. During the clustering process， it is found that the unascertained clustering algorithm has good performance in efficiency and accuracy.

Keywords： patent quality evaluation; unascertained clustering; patent data analysis; evaluation model building; dataset clustering; contrast experiment

0  引  言

随着科学技术的进步与发展，人们对知识产权的重视程度越来越高[1]。根据2018年12月3日世界知识产权组织发布的《2018年世界知识产权指标》显示，我国在全球专利、商标和工业品外观设计等专利申请量上均位居第一。专利总数越多，质量就会参差不齐，大量的低质量专利反而会为社会带来许多弊端。

国内目前对专利质量评价的研究中，有学者研究显示专利引证指标和专利权利范围指标所占权重最高，其次是专利地域范围指标、技术应用范围和科学关联度指标[2]。在对具体领域的专利质量进行分析的时候，包英群等学者遵循普适性、简洁性和可行性的原则选择出针对液晶显示产业的质量评价指标为专利引证指标和同族专利數指标[3]。……