Article: Personalised recommendation algorithm and platform design of engineering ethics case teaching resources incorporating big data analysis technology Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.91 pp.1 - 22 Abstract: As engineering ethics teaching faces the contradiction between overload of case resources and increasing demand for individualisation, this study aims to develop a system that can provide accurate and explainable recommendations. By constructing the KGDP-RM model, which integrates knowledge graphs with differential privacy protection, we employ graph neural networks to learn case semantic association, and attention mechanism and Bayesian network are introduced to realise dynamic preference capture and causal explanation. The accuracy rate of the model on the EdNet dataset reaches 0.821, the F1-value is 0.813, representing an improvement of 0.038 over the best-performing baseline. Its performance degrades by only 0.024 in a 10% noise environment. The user relevance and satisfaction scores reach 4.3 and 4.1 points respectively. In addition, the ablation test confirms that the knowledge graph module contributes the most. Overall, the multi-technology integration framework proposed in this study effectively solves the problems of sparsity, privacy security and interpretability in teaching resource recommendation. Inderscience Publishers - linking academia, business and industry through research

Title: Personalised recommendation algorithm and platform design of engineering ethics case teaching resources incorporating big data analysis technology

Authors: Weifeng Liang

Addresses: School of Marxism, Guangdong University of Science and Technology, Dongguan, 523808, Guangdong, China

Abstract: As engineering ethics teaching faces the contradiction between overload of case resources and increasing demand for individualisation, this study aims to develop a system that can provide accurate and explainable recommendations. By constructing the KGDP-RM model, which integrates knowledge graphs with differential privacy protection, we employ graph neural networks to learn case semantic association, and attention mechanism and Bayesian network are introduced to realise dynamic preference capture and causal explanation. The accuracy rate of the model on the EdNet dataset reaches 0.821, the F1-value is 0.813, representing an improvement of 0.038 over the best-performing baseline. Its performance degrades by only 0.024 in a 10% noise environment. The user relevance and satisfaction scores reach 4.3 and 4.1 points respectively. In addition, the ablation test confirms that the knowledge graph module contributes the most. Overall, the multi-technology integration framework proposed in this study effectively solves the problems of sparsity, privacy security and interpretability in teaching resource recommendation.

Keywords: engineering ethics case teaching; personalised recommendations; knowledge graph; differential privacy; interpretability.

DOI: 10.1504/IJICT.2026.155781

International Journal of Information and Communication Technology, 2026 Vol.27 No.91, pp.1 - 22

Received: 03 Feb 2026
Accepted: 05 Jun 2026
Published online: 13 Aug 2026 *

Title: Personalised recommendation algorithm and platform design of engineering ethics case teaching resources incorporating big data analysis technology

Keep up-to-date