突破单维局限:多层网络分析框架下的学习者交互解构  

Beyond Unidimensional Analysis:Deconstructing Learner Interactions Through a Multilayer Network Framework

作　　者：柴唤友 胡天慧 粟柱[1] 郭方琪 刘智[1] Chai Huanyou;Hu Tianhui;Su Zhu;Guo Fangqi;Liu Zhi(Faculty of Artificial Intelligence in Education,China Central Normal University,Wuhan Hubei 430079;School of Computers,Hubei University of Education,Wuhan Hubei 430205)

机构地区：[1]华中师范大学人工智能教育学部,湖北武汉430079 [2]湖北第二师范学院计算机学院,湖北武汉430205

出　　处：《远程教育杂志》2025年第2期85-93,共9页Journal of Distance Education

基　　金：国家自然科学基金项目“融合机器学习的协作问题解决能力归因推理方法研究”(项目编号:62407019);教育部人文社科规划基金项目“在线学习者多层交互网络的微观模式识别及其演化规律研究”(项目编号:24YJC880049);湖北省教育科学规划基金项目“基于多模态数据的自主学习投入智能测评与干预研究”(项目编号:2024GB136)的研究成果。

摘　　要：学习者交互是在线学习中学习者协作知识构建的发生基础和形成过程,对其开展分析有助于理解在线学习发生方式及成效,但传统社会网络分析难以综合解析学习者交互中更多维度或方面的信息。多层网络分析能够有效应对该不足,为学习者交互分析提供新方法。多层网络分析旨在研究由多个层次构成的复杂网络系统,以及不同层次系统之间的相互作用和依赖关系,可提供关于多层网络数据的个体、群体和总体水平的表征。与社会网络分析相比,多层网络分析具备驱动多维数据全面采集、综合分析和优质解读三大优势。展望未来,多层网络分析在学习者交互研究中的应用方向包括识别有影响力的节点或连边、量化局部水平的网络关系以及开展多时序联合建模,以驱动理解学习者个体作用、捕捉群体结构动态和探析总体演变过程。Learner interaction is fundamental to the process of collaborative knowledge construction in online learning.Analyzing these interactions is critical for understanding how online learning unfolds and evaluating its effectiveness.However,traditional social network analysis(SNA)falls short in capturing the multidimensional and dynamic nature of learner interactions.Multilayer network analysis(MNA)offers a promising alternative by modeling complex systems composed of multiple interconnected layers and examining the interactions and dependencies across them.This approach enables a comprehensive analysis of learner interaction data at the individual,group,and system-wide levels.Compared to SNA,MNA presents three distinct advantages:it facilitates the systematic collection of multidimensional data,supports integrative and nuanced analysis,and enhances the interpretability of networked learning processes.As such,MNA holds considerable potential for advancing research on learner interactions.Promising directions for future inquiry include identifying influential nodes and edges,quantifying local-level network dynamics,and modeling interactions across multiple temporal dimensions.These applications contribute to a deeper understanding of individual learner roles,evolving group structures,and the overall development of online learning networks.

关 键 词：在线教育 学习者交互 社会网络分析 多层网络分析 复杂网络 

分 类 号：G420[文化科学—课程与教学论]