摘要
随着中国经济转型升级步伐的加快,推进全国各地知识产权建设已成为创新驱动发展战略的重要议题。虽然我国已成为知识产权大国,但却非知识产权强国,因此构建完善的高价值专利识别方法将有助于专利价值精准分类,从而更有效地完善专利布局、提升研发效益、制定知识产权战略等。选取珠海市高新技术上市企业发明专利数据,提出以专利形成的全周期过程构建高价值专利识别指标,即高水平技术研发类指标、高质量申请确权类指标、高回报转化运用类指标,基于支持向量机、神经网络、自适应增强这三类机器学习法搭建高价值专利识别模型,并进行实证分析。通过完善高价值专利识别体系,助力企业和决策部门高价值专利识别工作的开展。
With the acceleration of China's economic transformation and upgrading,promoting intellectual property construction across the country has become an important issue in the innovation driven development strategy.Although China has become a major intellectual property country,it is not a strong intellectual property country.Therefore,building a comprehensive method for identifying high-value patents will help to accurately classify patent values,thereby more effectively improving patent layout,enhancing research and development efficiency,and formulating intellectual property strategies.This study selects the invention patent data of high-tech enterprises in Zhuhai City and proposes to construct high-value patent recognition indicators based on the full cycle process of patent formation,namely high-level technology research and development indicators,high-quality application confirmation indicators,and high return conversion application indicators.Based on three types of machine learning methods,namely support vector machine,neural network,and adaptive enhancement,a high-value patent recognition model is constructed and empirical analysis is conducted.This study aims to improve the high-value patent identification system and assist enterprises and decision-making departments in carrying out high-value patent identification work.
作者
夏芸
魏田苡薇
洪楷宣
马硕
XIA Yun;WEI Tianyiwei;HONG Kaixuan;MA Shuo(International Business School,Jinan University,Zhuhai 519070,China)
出处
《科学与管理》
2024年第4期1-9,F0003,共10页
Science and Management
基金
珠海市哲学社会科学规划课题“知识产权强市战略下珠海市高新技术企业高价值专利判别、测度与培育路径研究”(2023YBB030)。
关键词
高价值专利
机器学习
支持向量机
神经网络
自适应增强
high-value patents
machine learning
support vector machine
neural network
adaptive boosting