Purpose

This study aims to find technologically important patent identification methods and indicators early and efficiently to grasp the technical qualitative level of patents, which are output indicators of research and development (R&D) results.

Design/methodology/approach

This paper reports on two methods for distinguishing important patents and the indicators obtained from those methods. One of the discrimination methods is Heckman's two-step estimation procedure. The second method is to find the centrality of each patent by network analysis of the citation relationship between publications and to find the importance from the magnitude of the centrality value.

Findings

In Heckman's analysis, the number of citations within three years after publication and the applicant's right acquisition/maintenance motivation index had positive effects on patent importance. The discriminative indicators of important patents by network analysis were degree centrality, mediation centrality, proximity centrality and transit values in the aggregated subnetworks. These two analytical methods are in a relationship that can complement each other's shortcomings. To efficiently evaluate the qualitative importance of patents, it is recommended to use these two methods together.

Research limitations/implications

The indicators of important technical patents might change depending on the technical field. Future studies can apply this research to multiple technical fields to improve robustness and to construct an algorithm that can efficiently evaluate the quality of patents.

Practical implications

This study's results can be useful for grasping the patent position of the company or competitors numerically and for quantitatively evaluating the quality of R&D activities. Furthermore, it is possible to streamline the routine for an exploratory search of a huge number of patents. For example, it could be useful for detecting changes in the paradigm of specific technical knowledge, evolving the genealogy of technical knowledge and creating patent maps for new R&D. These methods greatly increase the effectiveness of technical knowledge information, which is the basis of R&D. In addition, the results of this study can help in evaluating patented assets.

Social implications

This study confirmed the development process of technical knowledge. It is a fact that sharing, sympathy and mutual trust for technical issues and technical values are created among professional engineers and researchers inside and outside the organization, and their preferences and interactions develop and expand technical knowledge. Understanding the process of development and the evolution of this technical knowledge gives hints, such as expanding the discretionary power of engineers and researchers regarding corporate secrets, or reviewing the balance between control and independence, to solve Japanese management problems, which are often closed and monetized in R&D activities.

Originality/value

This study presents a scoring of the technical significance of patents by combining the two analytical methods. In addition, there are proposals as a method for detecting changes in the genealogy and paradigm of technical knowledge. As an analysis method, it is a new proposal that has never existed before.

中文翻译：

---

重要专利认定方法及指标

目的

本研究旨在及早、高效地发现技术上重要的专利识别方法和指标，以掌握专利的技术定性水平，即研发（R&D）成果的产出指标。

设计/方法/方法

本文报告了区分重要专利的两种方法以及从这些方法中获得的指标。鉴别方法之一是赫克曼的两步估计程序。第二种方法是通过对出版物之间的引用关系进行网络分析，找出每个专利的中心度，并从中心度值的大小上找出重要性。

发现

在赫克曼的分析中，发表后三年内的引用次数和申请人的权利获取/维持动机指数对专利重要性有正向影响。网络分析对重要专利的判别指标为聚合子网络中的度中心性、中介中心性、邻近中心性和中转值。这两种分析方法是可以互补的关系。为了有效地评估专利的定性重要性，建议将这两种方法结合使用。

研究限制/影响

重要技术专利的指标可能因技术领域而异。未来的研究可以将这项研究应用于多个技术领域，以提高鲁棒性并构建能够有效评估专利质量的算法。

实际影响

本研究的结果可用于从数字上掌握公司或竞争对手的专利地位，并用于定量评估研发活动的质量。此外，可以简化对大量专利进行探索性检索的程序。例如，它可以用于检测特定技术知识范式的变化、发展技术知识的谱系以及为新的研发创建专利地图。这些方法大大提高了技术知识信息的有效性，这是研发的基础。此外，本研究的结果有助于评估专利资产。

社会影响

本研究证实了技术知识的发展过程。事实上，组织内外的专业工程师和研究人员对技术问题和技术价值产生了共享、同情和相互信任，他们的偏好和互动发展和扩展了技术知识。了解这些技术知识的发展过程和演变过程提供了一些提示，例如扩大工程师和研究人员对公司机密的自由裁量权，或审查控制与独立之间的平衡，以解决日本的管理问题，这些问题通常是封闭和货币化的在研发活动中。

原创性/价值

本研究通过结合两种分析方法对专利的技术意义进行评分。此外，还提出了作为检测技术知识的谱系和范式变化的方法的建议。作为一种分析方法，它是一种前所未有的新提案。