Scholarlink

<p>With increasing attention being paid to climate change adaptation (CCA) following the Paris Agreement, countries have been striving to develop related technologies. Consequently, the cooperative patent classification (CPC) system introduced a new category for CCA technologies, Y02A, in 2018. Identifying the excellence of patents in advance is crucial for developing high-quality CCA technologies; however, a comprehensive framework is lacking. Therefore, this study constructed a patent valuation model based on machine learning focusing on Y02A patents registered between 2019 and 2023. The dataset comprises 6,705 USPTO patents classified under Y02A. By utilizing CPC codes and claims and using the Korea Invention Promotion Association’s system used to measure, analyze, and rate patent technology (SMART5) evaluation features, we developed a machine learning evaluation model for CCA patents that combines CPC codes and claims. The results show that the combined model based on SVM achieved higher accuracy (68.51%) and F1-score (weighted averaged 66.84%) than the individual models using only CPC codes or claims. This study is expected to contribute to various fields such as business management and policy by enabling the early detection of CCA-related technologies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The identification of valuable climate change adaptation patents based on machine learning approaches

  • 期刊论文

  • 发表日期：

    2025年12月10日

查看全文
  • O. Kwon,
  • S. Nam

摘要

With increasing attention being paid to climate change adaptation (CCA) following the Paris Agreement, countries have been striving to develop related technologies. Consequently, the cooperative patent classification (CPC) system introduced a new category for CCA technologies, Y02A, in 2018. Identifying the excellence of patents in advance is crucial for developing high-quality CCA technologies; however, a comprehensive framework is lacking. Therefore, this study constructed a patent valuation model based on machine learning focusing on Y02A patents registered between 2019 and 2023. The dataset comprises 6,705 USPTO patents classified under Y02A. By utilizing CPC codes and claims and using the Korea Invention Promotion Association’s system used to measure, analyze, and rate patent technology (SMART5) evaluation features, we developed a machine learning evaluation model for CCA patents that combines CPC codes and claims. The results show that the combined model based on SVM achieved higher accuracy (68.51%) and F1-score (weighted averaged 66.84%) than the individual models using only CPC codes or claims. This study is expected to contribute to various fields such as business management and policy by enabling the early detection of CCA-related technologies.