Development and Application of Digital Teacher Portrait Empowered by Large Multimodal Model
Digital portraits have emerged as a crucial lever to facilitate high-quality progress in teachers'professional development.This research aims to address the problems in teacher digital portrait technology,including structural mismatches in data collection,imbalances between model generalization and depth,temporal lags in model training and upgrading,as well as the lack of accompanying application services.Specifically,the research proposes a teacher portrait large multimodal model with four key capabilities:Multimodal cognition,associative reasoning,incremental learning,and accompanying feedback.By designing a multi-agent system,this model empowers the integrated services to achieve both precise depictions of teacher capabilities and competency portraits multidimensionally with the evidence-based professional development of teachers through human-machine collaborations.With an example of intelligent assessment of teachers'digital literacy,this research elucidates the practical pathways and effectiveness of applying the teacher portrait large multimodal system through dialog-based intelligent assessment,hierarchical portrait analysis,and accompanying literacy development.
digital portraitlarge multimodal modelteacher professional developmentmulti-agent system
万方数据
维普
