The artificial intelligence patent dataset (AIPD) 2023 update

被引:1
作者
Pairolero, Nicholas A. [1 ]
Giczy, Alexander V. [1 ,2 ]
Torres, Gerard [3 ]
Erana, Tisa Islam [3 ]
Finlayson, Mark A. [1 ,3 ]
Toole, Andrew A. [1 ]
机构
[1] US Patent & Trademark Off, Alexandria, VA 22314 USA
[2] Addx Corp, Alexandria, VA USA
[3] Florida Int Univ, Miami, FL USA
关键词
Patent; Patent landscape; Artificial intelligence; AI; AI patent dataset; AIPD; O31; O34; C45; L86;
D O I
10.1007/s10961-025-10189-8
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
The 2023 update to the Artificial Intelligence Patent Dataset (AIPD) extends the original AIPD to all United States Patent and Trademark Office (USPTO) patent documents (i.e., patents and pre-grant publications, or PGPubs) published through 2023, while incorporating an improved patent landscaping methodology to identify AI within patents and PGPubs. This new approach substitutes BERT for Patents for the Word2Vec embeddings used previously, and uses active learning to incorporate additional training data closer to the "decision boundary" between AI and not-AI to help improve predictions. We show that this new approach achieves substantially better performance than the original methodology on a set of patent documents where the two methods disagreed-on this set, the AIPD 2023 achieved precision of 68.18 percent and recall of 78.95 percent, while the original AIPD achieved 50 percent and 21.05 percent, respectively. To help researchers, practitioners, and policy-makers better understand the determinants and impacts of AI invention, we have made the AIPD 2023 publicly available on the USPTO's economic research web page.
引用
收藏
页数:24
相关论文
共 41 条
[1]   Automated patent landscaping [J].
Abood, Aaron ;
Feltenberger, Dave .
ARTIFICIAL INTELLIGENCE AND LAW, 2018, 26 (02) :103-125
[2]  
[Anonymous], 2019, WORLD INTELLECTUAL P
[3]   Winner takes all? Tech clusters, population centers, and the spatial transformation of US invention [J].
Chattergoon, B. ;
Kerr, W. R. .
RESEARCH POLICY, 2022, 51 (02)
[4]   Is your search query well-formed? A natural query understanding for patent prior art search [J].
Chikkamath, Renukswamy ;
Rastogi, Deepak ;
Maan, Mahesh ;
Endres, Markus .
WORLD PATENT INFORMATION, 2024, 76
[5]   Deep learning learning for patent landscaping using transformer and graph embedding [J].
Choi, Seokkyu ;
Lee, Hyeonju ;
Park, Eunjeong ;
Choi, Sungchul .
TECHNOLOGICAL FORECASTING AND SOCIAL CHANGE, 2022, 175
[6]   Public support for research in artificial intelligence: a descriptive study of US Department of Defense SBIR Projects [J].
Chowdhury, Farhat ;
Link, Albert N. ;
van Hasselt, Martijn .
JOURNAL OF TECHNOLOGY TRANSFER, 2022, 47 (03) :762-774
[7]  
Cockburn IainM., 2019, The economics of artificial intelligence: An agenda, P115, DOI DOI 10.7208/CHICAGO/9780226613475.003.0004
[8]  
Devlin J, 2019, Arxiv, DOI [arXiv:1810.04805, DOI 10.48550/ARXIV.1810.04805]
[9]  
Erana TI, 2024, Arxiv, DOI arXiv:2407.08001
[10]   Patent eligibility after Alice: Evidence from USPTO patent examination [J].
Frumkin, Jesse ;
Pairolero, Nicholas A. ;
Tesfayesus, Asrat ;
Toole, Andrew A. .
JOURNAL OF ECONOMICS & MANAGEMENT STRATEGY, 2024, 33 (03) :748-769