Title: Back propagation neural network in artificial intelligence for intellectual property protection in the automotive industry

Authors: Xiaolong Liang; Kewei Ji; Tiantian Qu; Huiting Wang; Qikun Ao

Addresses: College of Intellectual Property, Hubei University of Automotive Technology, Shiyan, Wuhan, Hubei, China ' Design School, Xi'an Jiaotong-Liverpool University, Suzhou, Jiangsu, China ' School of Automotive Business, Hubei University of Automotive Technology, Shiyan, Wuhan, Hubei, China ' School of Automotive Business, Hubei University of Automotive Technology, Shiyan, Wuhan, Hubei, China ' Department of Ideological and Political Education, Yangjiang Polytechnic, Yangjiang, Guangdong, China; School of Law, Dong-A University, Busan, South Korea

Abstract: With the rapid development of the New Energy Vehicles (NEVs) industry, how to scientifically evaluate patent value and improve intellectual property protection efficiency has become an important issue. This study focuses on intellectual property protection in the automotive industry, using NEVs as the subject, and constructs a patent value evaluation index system. The eXtreme Gradient Boosting (XGBoost) algorithm is employed to select feature indicators, and a high-precision patent value evaluation model, XGBoost and SAM-Based Back Propagation Neural Network (XSA-BPNN), is developed. The experimental results show that the XGBoost algorithm performs the best in feature selection. Its Mean Squared Error (MSE) is 0.127, Mean Absolute Error (MAE) is 0.198, explained variance is 0.438 and R² is 0.413. These values significantly outperform the Random Forest and Gradient Boosting Tree algorithms. Finally, the XSA-BPNN model achieves an average absolute error of only 0.0484 in patent value evaluation, significantly outperforming the comparison models. This indicates that the proposed XSA-BPNN model enhances the accuracy of patent value evaluation by precisely capturing and utilising key features. This study provides technical support for optimising patent transactions and management processes, as well as improving the scientific approach to intellectual property protection.

Keywords: automotive industry; patent value evaluation; XGBoost; BPNN; intellectual property protection.

DOI: 10.1504/IJCAT.2025.149869

International Journal of Computer Applications in Technology, 2025 Vol.77 No.1/2, pp.117 - 132

Received: 06 Mar 2025
Accepted: 26 Jun 2025
Published online: 14 Nov 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article