Knowledge graph as an adaptive cognitive scaffolding: enhancing learning outcomes in CLIL-based international trade...

Driven by the digital transformation of higher education, innovative instructional methodologies in Content and Language Integrated Learning (CLIL) are pivotal for developing interdisciplinary foreign language professionals. To address the dual cognitive load inherent in CLIL environments, this study proposes a Knowledge Graph (KG)-driven adaptive scaffolding framework and evaluates its effectiveness within an Outcome-Based Education (OBE) paradigm. Adopting an educational design research methodology, we implemented a 16-week intervention targeting 51 undergraduate English majors at a Chinese application-oriented university. To assess the model's efficacy in promoting disciplinary and linguistic gains, the study utilized a multi-dimensional evaluation matrix and a normalized Course Objective (CO) attainment matrix aligned with the professional accreditation standards. Quantitative analyses revealed substantial improvements in domain-specific literacy (overall CO attainment = 0.78), with particularly notable performance in disciplinary knowledge (CO1 = 0.81) and language proficiency (CO2 = 0.82). However, a "knowledge-action gap" persisted in the development of practical skills (CO3 = 0.73). These findings suggest that while KG demonstrates exceptional strength in conceptual scaffolding, additional immersive simulations are required to cultivate high-order procedural knowledge. By addressing these cognitive and practical dimensions, the study provides a scalable framework for integrating KG-driven adaptive scaffolding into New Liberal Arts curricula, while also identifying the limits of its scaffolding approach in fostering practical expertise.