基于跨模态理解与重构的适应性数字教育资源:模型构建与实践框架

Adaptive Digital Educational Resources Based on Cross-Modal Understanding and Reconstruction:Model Construction and Practical Framework

数字教育资源在推进教育高质量发展的过程中扮演着重要角色。人工智能时代,如何让数字教育资源从之前的“让人适应”变为能“主动适应人”,是数字教育资源建设亟待解决的关键问题。随着人工智能生成内容(AIGC)应用场景的持续拓展,跨模态理解与重构技术为破解数字教育资源的适应性问题提供了可行的技术支持。具体而言,就是主张“人—境—机—物”多主体协同,利用跨模态理解与重构技术对复杂教育情境中的多模态教学行为和多模态数字教育资源进行智能挖掘和耦合计算,实现适用于人机协同环境的教育资源推荐,挖掘数字教育资源服务的供需适配关系,达成数字教育资源生成链的自我调度,由此即可构建起适应性数字教育资源模型。在实践推进层面,应构建起包括数据采集层、智能分析层、规则约束层、应用服务层四个层次的适应性数字教育资源实践框架,以期在对多模态教育资源和师生行为进行跨模态理解的基础上,切实推进对数字教育资源的智能管理和个性化服务。

Digital educational resources play an important role in the process of promoting high-quality development of education.In the era of artificial intelligence,how to make digital educational resources change from“letting human adapt”to“actively adapting to human”is the key issue to be solved urgently in the construction of digital educational resources.As the application scenarios of artificial intelligence-generated content continue to expand,cross-modal understanding and reconstruction technologies provide feasible technical support for addressing the adaptability of digital educational resources.Specifically,it advocates multi-disciplinary collaboration among humans,educational contexts,computers and educational resources,and uses cross-modal understanding and reconstruction technology to carry out intelligent mining and coupling calculation of multi-modal teaching behaviors and multi-modal digital educational resources in complex educational situations,so as to realize the recommendation of educational resources suitable for man-machine collaborative environment,explore the supply-demand adaptation of digital educational resources services,and realize the self-scheduling of digital educational resources generation chain,so that we can build an adaptive digital educational resources model.At a practical level,a practical framework of adaptive digital educational resources should be constructed,which includes data collection layer,intelligent analysis layer,rule constraint layer and application service layer,to effectively promote the intelligent management and personalized service of digital educational resources on the basis of cross-modal understanding of multi-modal educational resources as well as behaviors of teachers and students.