Research on the Construction of a Dynamic Evaluation Model for Vocational Education Quality Based on Multimodal Learning Analytics

Abstract

Driven by globalization and educational informatization, new models of vocational education have emerged, posing significant challenges to traditional mechanisms of vocational education quality assessment and certification. Existing evaluation frameworks are constrained by two critical limitations: data silos and timeliness lag, which are reflected in issues such as single-source data dependence, static and retrospective assessment processes, difficulties in establishing cross-border data trust, and restricted circulation of certification outcomes. From an international perspective, this study constructs a dynamic evaluation model for vocational education quality by integrating blockchain technology with multimodal learning analytics theory. Leveraging the inherent characteristics of blockchain, the model establishes a cross-border and cross-institutional trusted network for the authentication and circulation of learning achievements, enabling continuous tracking of learners' cross-border learning trajectories. By incorporating multimodal learning analytics techniques, the model conducts in-depth mining of multi-source learning data and integrates sentiment recognition to achieve comprehensive, fine-grained, and real-time analysis of learning processes. Ultimately, through smart contract-enabled data aggregation, the model generates traceable, quantifiable, and dynamically updated learner competency profiles and evaluation reports, providing effective support and practical pathways for quality assessment and certification in internationalized vocational education.