在线教与学集体智慧的有效利用:学习分析的视角与架构

Using Collective Intelligence to Support Online Learning and Teaching: A Learning Analytics Perspective

在各类在线学习平台或社会性学习工具的支持下,越来越多的个体学习活动被有效组织和连接起来,共同创造出一种名为集体智慧的新知识。集体智慧由个体行为以显性或隐性的方式汇聚而成,体现为一种共同的认知或行为的状态或趋势,并在社会情境下整体输出,发挥成效。这种知识被证实在解决复杂问题上要优于专家智慧。为了探讨如何利用学习分析技术促进集体智慧的获取与应用,进而有效支持在线教与学的教学决策,本研究在梳理现有学习分析模型操作要素的基础上,从目标确定、数据收集、分析过程和结果应用四方面提出了相应的理论框架。集体智慧的学习分析目标可以依据教师和学生的兴趣和关注点加以确立,数据分析可分为社会网络分析、话语分析、内容分析、性格分析和情境分析五大类,具体操作时还可从时间维度再细分为过去、现在和未来三种,相应的服务也可划分为信息级和洞悉级两个层次,但都应考虑分析结果呈现的方式以及应用策略。本研究最后通过应用案例对所提框架加以诠释,并讨论了未来需要开展的工作,包括定义各种服务于不同分析目标的OIM关系链、设计和开发解读和应用相关分析结果的工具、开展试验研究加以验证等。

With the support of online learning platforms and social learning tools,individual learning activities can be effectively organized and linked collectively to generate a new knowledge named collective intelligence. Collective intelligence is behaviors performed explicitly or implicitly by individuals,reflecting a common state or trend,and acted as an aggregate in social context with favorable integral outcome. This study discussed the use of learning analytics to support the generation of collective intelligence for online learning and teaching support. Combing existing representative models of learning analytics,this study proposed a theoretical framework from four aspects,including objectives setting,data collection,analysis processing,and application of results,for the practice of collective intelligence and learning analytics to facilitate informed pedagogical decision making. Objectives can be set up according to the interests of teachers and learners,and learning data can come from social network,discourse,content,disposition and context. Three orientations（ past,present and future） and two levels（ information and insights） of collective intelligence and learning analytics usages were suggested in the framework,followed by the methods for result presentation and the strategies for result application. An application demo was given to further demonstrate the proposed framework. At the end of this study,future works were discussed,including defining various OIM triples for different objectives,developing supporting tools for easily interpreting analytical results,and evaluating the effect of the framework through experimental studies.