数据驱动的多维融合在线协作会话可计算模型及应用研究  

Research on the Computable Model and Application of Data-driven Multi-dimensional Fussed Online Collaborative Discourse

作　　者：吴林静[1] 高喻 涂凤娇 马鑫倩 刘清堂[1] WU Lin-jing;GAO Yu;TU Feng-jiao;MA Xin-qian;LIU Qing-tang(Faculty of Artificial Intelligent in Education,Central China Normal University,Wuhan,Hubei,China 430079)

机构地区：[1]华中师范大学人工智能教育学部,湖北武汉430079

出　　处：《现代教育技术》2023年第4期101-110,共10页Modern Educational Technology

基　　金：国家自然科学基金项目“数据驱动的在线学习协作会话过程监测与干预机制研究”(项目编号:72174070);华中师范大学“人工智能+教育”教学创新研究项目“在线协作会话中不同的教师干预策略对学习者认知加工过程的影响研究”(项目编号:2022XY023)、华中师范大学信息化与基础教育均衡发展省部共建协同创新中心重点项目“中小学教师信息化教学胜任力指标构建及智能测评”(项目编号:xtzd2021-008)的阶段性研究成果。

摘　　要：在线协作会话是在线学习活动的重要组成部分,也是促进学习者进行协同知识建构的重要手段,因此越来越受关注。目前,由于分析框架和分析方法不够完善,导致当前在线协作会话存在分析维度单一、自动化程度较低等问题。为此,文章首先提出了一种数据驱动的多维融合在线协作会话分析框架和可计算模型,然后提出了协作水平系数和协作投入倾向两个核心指标对会话状态进行整体评估,最后进行了案例分析。案例分析的结果表明:协作水平系数能够准确表征学习者在在线协作会话中的协作状态;在线协作会话可计算模型可以直观地呈现学习者在线协作会话过程中的核心特征和学习者的个体差异性;学习者的在线协作会话模式包括领导型、思想型、中庸型和边缘型四种典型类型。文章通过研究验证了模型的有效性和其对于在线协作会话自动分析与评估的实践价值,期望能够为在线协作会话的自动化分析提供参考。As an important part of online learning activities,online collaborative discourse is also an important method to promote learners to carry out collaborative knowledge construction,which has attracted more and more attention.However,due to not enough perfect analysis framework and analysis methods,the current online collaborative discourse exists the problems such as single analysis dimension and low degree of automation.Therefore,this paper firstly proposed a computable model of data-driven and multi-dimensional fused online collaborative discourse,and then put forward to evaluate the overall state of of collaborative discourse by tow core indicators of collaboration level coefficient and collaboration engagement tendency,and finally made a case study.The results of the case study showed that collaboration level coefficient could accurately represent learners’collaborative state in collaborative discourse,and the computable model of online collaborative discourse could intuitively present learners’core characteristics and individual differences in the process of online collaborative discourse.Meanwhile,learners’online collaborative discourse model included four types of leadership type,thought type,moderation type and marginal type.Through research,the effectiveness of the model and its practical value for automatic analysis and evaluation of online collaborative discourse were verified in this paper,which was expected to provide reference for the automatic analysis of online collaborative discourse.

关 键 词：在线协作会话 数据驱动 可计算模型 协作水平系数 

分 类 号：G40-057[文化科学—教育学原理]