Article: Analysis of social English learning behaviour based on federated knowledge graph and privacy preservation Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.26 pp.68 - 86 Abstract: Social learning platforms provide a dynamic interactive and individualised learning experience as artificial intelligence in education is increasingly used. Though they have data privacy and AI security issues, these services gather a great amount user data. Releasing user data such as voice calls, text messages, social connections, and learning preferences could cause privacy issues. This paper presents FL-KGPrivacyEdu, a federated knowledge graph model combining federated learning (FL) and knowledge graph (KG) to address the problem of behavioural modelling and privacy preservation in social English learning. The methodology examines social learning across platforms and terminals while safeguarding data. Experimental validation demonstrates that the model is successful and superior in social English learning behaviour analysis in accuracy, recall, and F1 score. This study also looks at how model convergence, a major model training reference, is affected by learning rate modification in worldwide rounds. Inderscience Publishers - linking academia, business and industry through research

Title: Analysis of social English learning behaviour based on federated knowledge graph and privacy preservation

Authors: Gang Shen; Tao Feng

Addresses: School of Foreign Studies, Suqian University, Suqian 223800, China ' School of Foreign Studies, Suqian University, Suqian 223800, China

Abstract: Social learning platforms provide a dynamic interactive and individualised learning experience as artificial intelligence in education is increasingly used. Though they have data privacy and AI security issues, these services gather a great amount user data. Releasing user data such as voice calls, text messages, social connections, and learning preferences could cause privacy issues. This paper presents FL-KGPrivacyEdu, a federated knowledge graph model combining federated learning (FL) and knowledge graph (KG) to address the problem of behavioural modelling and privacy preservation in social English learning. The methodology examines social learning across platforms and terminals while safeguarding data. Experimental validation demonstrates that the model is successful and superior in social English learning behaviour analysis in accuracy, recall, and F1 score. This study also looks at how model convergence, a major model training reference, is affected by learning rate modification in worldwide rounds.

Keywords: federated learning; FL; knowledge graph; KG; federated knowledge graph; data privacy protection; AI security.

DOI: 10.1504/IJICT.2025.147468

International Journal of Information and Communication Technology, 2025 Vol.26 No.26, pp.68 - 86

Received: 14 May 2025
Accepted: 25 May 2025
Published online: 16 Jul 2025 *

Title: Analysis of social English learning behaviour based on federated knowledge graph and privacy preservation

Keep up-to-date