| 注册
首页|期刊导航|现代情报|基于维度复合和可解释性说明的高价值专利评估体系构建研究

基于维度复合和可解释性说明的高价值专利评估体系构建研究

窦路遥 魏凤 邓阿妹 周洪

现代情报2025,Vol.45Issue(11):30-42,13.
现代情报2025,Vol.45Issue(11):30-42,13.DOI:10.3969/j.issn.1008-0821.2025.11.004

基于维度复合和可解释性说明的高价值专利评估体系构建研究

Research on the Construction of a High-Value Patent Evaluation System Based on Multi-dimensional Integration and Explainable Interpretation

窦路遥 1魏凤 2邓阿妹 3周洪3

作者信息

  • 1. 中国科学院武汉文献情报中心,湖北 武汉 430071||中国科学院大学经济与管理学院信息资源管理系,北京 100190
  • 2. 中国科学院武汉文献情报中心,湖北 武汉 430071||中国科学院大学经济与管理学院信息资源管理系,北京 100190||科技大数据湖北省重点实验室,湖北 武汉 430071
  • 3. 中国科学院武汉文献情报中心,湖北 武汉 430071||科技大数据湖北省重点实验室,湖北 武汉 430071
  • 折叠

摘要

Abstract

[Purpose/Significance]Establishing a comprehensive,applicable,and credible evaluation system for high-value patents is critical for enhancing national intellectual property competitiveness,optimizing the national innova-tion patent layout,and strengthening strategic scientific and technological capabilities.[Process/Method]Based on all existing high-value patent evaluation systems,this study aggregated and organized evaluation indicators from various sys-tems.By constructing a full-indicator map,a dimension-overlap map,and single-dimension indicator maps,a novel high-value patent evaluation system was built across five dimensions:economic,market,legal,strategic,and technological.Subsequently,the entropy weight-catastrophe progression method was employed to calculate the value of the constructed evaluation system over time.High-value patents were then identified based on the Pareto principle.Finally,the feasibility of the system was validated using a deep learning model,and indicator importance was interpreted based on ranking impor-tance derived from combinations of indicators.[Result/Conclusion]An empirical study involving 65 062 patents in the lithium-ion battery field reveals that the newly constructed high-value patent evaluation system is more comprehensive at both the indicator and dimensional levels.The system's feasibility is confirmed through deep learning validation.Key indi-cators influencing high-value patents include emerging industry classification,protection scope,high-tech attributes,technology stability,and the number of inventors.

关键词

高价值专利/评估体系构建/知识产权/深度学习/锂离子电池

Key words

high-value patents/evaluation system construction/intellectual property/deep learning/lithium-ion battery

引用本文复制引用

窦路遥,魏凤,邓阿妹,周洪..基于维度复合和可解释性说明的高价值专利评估体系构建研究[J].现代情报,2025,45(11):30-42,13.

基金项目

国家电网公司总部科技项目"双碳目标下电力绿色低碳关键支撑技术评价方法和专利标准化研究"(项目编号No.1400-202340338A-1-1-ZN). (项目编号No.1400-202340338A-1-1-ZN)

现代情报

OA北大核心

1008-0821

访问量0
|
下载量0
段落导航相关论文