智能学习分组:从通用模型到大数据框架

Intelligent Learning Grouping:From Generic Models to Big Data Frameworks

学习分组是协作学习活动设计的首要阶段。随着学习场所的快速变化、多模态交互过程复杂性的增加,采用传统的随机分组、教师指派或学生自我选择等方法进行协作学习分组的效率十分低下。研究提出基于智能技术构建自适应的协作学习小组。首先,阐述了学习分组的价值,即构建合理的协作学习环境、兼顾学生的个体差异和促进教育资源优质公平;其次,总结了影响智能学习分组的因素,包括个体属性、小组学术与物理构成以及学习者与环境的交互;最后,描述了经典场景下智能学习分组的通用模型,并讨论了大数据背景下智能学习分组的前景与挑战。针对大数据驱动智能学习分组的稳定性问题,基于机器学习中的集成学习思想构建了大数据共识分组框架。此框架有望为人工智能促进未来规模化的个性化教育提供支持。

Learning grouping is the primary stage of designing collaborative learning activities.With the rapid changes in learning venues and the increasing complexity of multimodal interaction processes,the efficiency of collaborative learning grouping through traditional methods such as random grouping,teacher assignment,or student self-selection approaches is unsatisfactory.The study proposes the construction of adaptive collaborative learning groups based on intelligent technologies.First of all,this paper explains the value of learning grouping in terms of building a reasonable collaborative learning environment,taking into account students'individual differences and promoting quality and equity of educational resources.Then,this paper summarizes the factors that affect intelligent learning grouping,including individual attributes,group academic and physical composition,and learner-environment interaction.Finally,this paper describes a generic model of intelligent learning grouping in classic scenarios and discusses the prospects and challenges of intelligent learning grouping in the context of big data.A consensus grouping framework for big data is constructed based on the idea of integrated learning in machine learning to address the stability problem of big data-driven intelligent learning grouping.The framework is expected to provide support for artificial intelligence to promote large-scale personalized education in the future.