Early detection of valuable technologies: a BP neural network involving patent-based indicators

Dejian Yu and Zhaoping Yan ()
Additional contact information
Dejian Yu: Nanjing Audit University, School of Business
Zhaoping Yan: Nanjing University, School of Information Management

Scientometrics, 2025, vol. 130, issue 11, No 1, 5891 pages

Abstract: Abstract Predicting the patent value is an essential component of corporate strategic planning. We proposed a patent value index system consisting 17 features, utilizing forward citations as a proxy for patent value. We then constructed machine learning models to predict the patent value of 108,781 patents in the low emission vehicle (LEV) field. Finally, we evaluated the impact of different features on prediction performance and assess the accuracy of patent value classification. Our experiments show that BPNN outperforms five other baseline models on the test set. We identified 9 features that significantly affect the patent value prediction, including abstract length, patent novelty, 4-digit IPCs, CPCs, family count, claim length, assignee capabilities and scientific knowledge. By predicting patent value, this study can assist companies in identifying technological gaps and optimizing resource allocation.

Keywords: Patent value; Patent evaluation; Patent features; Back propagation neural network; Citation prediction (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11192-025-05471-3 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:scient:v:130:y:2025:i:11:d:10.1007_s11192-025-05471-3

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11192

DOI: 10.1007/s11192-025-05471-3

Access Statistics for this article

Scientometrics is currently edited by Wolfgang Glänzel

More articles in Scientometrics from Springer, Akadémiai Kiadó
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().