Personalized Textile Engineering English Teaching: An AI-Driven Adaptive Learning System for Optimizing Learning Outcomes

Abstract

To address the limitations of general-purpose artificial intelligence (AI) systems in supporting the professional context and personalized requirements of textile engineering English instruction, this study proposes an adaptive learning framework that integrates domain-specific knowledge graphs with dynamic learner modeling. A textile engineering English knowledge graph is constructed from multi-source corpora to organize semantic relationships among technical terminology, engineering processes, equipment, and standards, while deep knowledge tracing continuously models learner states and Q-learning dynamically optimizes individualized learning paths. Experimental evaluation demonstrates that the proposed system significantly improves terminology comprehension, achieving 86.4% recognition accuracy, technical document reading with 84.2% key information extraction accuracy, and cross-cultural engineering communication with an average score of 8.32/10, confirming the effectiveness of graph-guided adaptive instruction for English for Specific Purposes. Beyond educational applications, the integration of semantic knowledge representation, dynamic state estimation, and reinforcement learning provides a scalable framework for intelligent information processing and structured knowledge transmission, offering methodological implications for digital engineering systems and electromagnetic information networks that rely on efficient semantic organization and adaptive decision-making.