多元主体需求下的专利预警研究回顾与展望

Review and Prospect of Patent Early Warning Research Under Multi Subject Demand

以现有专利预警相关文献为研究对象,通过系统回顾与深度分析,对多元主体需求下的专利预警研究现状进行总结与梳理,分析其主要特征,并提出促进专利预警研究创新发展的改进思路与方向。研究发现:目前针对专利预警的概念内涵主要围绕专利威胁态势、专利竞争风险和专利侵权纠纷等视角展开探讨,关于专利预警机制建立的必要性、基本原则、关键作用、核心内容、主要措施等方面的研究构成了专利预警机制研究的理论体系,主要从预警主体、预警平台和预警指标等方面对专利预警体系的关键构成要素展开探索,专利预警的核心方法包括自然语言处理、机器学习、专利组合分析和专利地图分析,采集的信息资源包含专利信息及市场信息;然而在多源数据资源利用、预警模型构建及实时动态监测方面尚存在一定的不足。鉴于此,对后续研究提出两点展望:一是要进一步探索基于多源数据融合的专利预警实现机制;二是要深入探究基于扫描统计量的实时动态监测及预警方法,进而提高专利预警的有效性。

Taking the existing patent early warning literature as the research object, this paper summarizes the research status of patent early warning through systematic review and in-depth analysis, then analyzes the main characteristics of patent early warning, and proposes improved ideas and directions to promote the innovation and development of patent early warning research. It is found that the concept connotation of patent early warning is mainly discussed from the perspectives of patent threat situation, patent competition risk and patent infringement disputes;the research on the necessity, basic principles, key roles, core contents and main measures of establishing patent early warning mechanism constitutes the theoretical system of patent early warning mechanism research;key elements of patent early warning system are explored from the aspects of patent early warning subject, patent early warning platform and patent early warning indicators;core methods of patent early warning include natural language processing, machine learning, patent portfolio analysis and patent map analysis;and the collected information resources include patent information and market information. However, there are still some deficiencies in the utilization of multi-source data resources, the construction of early warning model and real-time dynamic monitoring. In view of this, two prospects are put forward for follow-up research: one is to explore the realization of patent early warning based on multi-source data fusion;the second is to explore real-time dynamic monitoring and early warning methods based on scanning statistics to improve the effectiveness of patent early warning.