【作 者】：
彭绍东
【关 键 词 】：
大数据； 网上学习；操作行为；活动行为；言语行为；挖掘方法模型
【栏 目】：
网络教育
【中文摘要】：
文章基于交互层次视角，利用大数据分析技术，对网上学习行为特征与规律研究的方法进行了探讨。所创建的“挖掘方法模型”，表达了针对三种挖掘对象（服务器日志、平台课程数据库、论坛帖子集）、沿着三条挖掘主线（操作行为挖掘流程线、活动行为挖掘流程线、言语行为挖掘流程线）、掘出三种行为特征（操作行为模式、活动行为规律、言语行为特征）的总框架。该模型有助于从网络服务器系统、教学平台系统和学习制品库中寻找隐含的网上学习规则与模式，对计算机支持的协作学习、混合式协作学习的深入研究具有重要支持作用。文中详解了模型内容，提炼了模型特征，并简介了模型应用的三个案例（“造访站点的路径排行挖掘”“BCL中协同认知加工行为的‘转化率’分析”“基于共词分析的语义网络特征挖掘”）。挖掘实证表明，该模型在研究中具有良好的可行性与有效性。
【英文摘要】：
From the perspective of interactive hierarchy and through big data analysis, a research method is introduced to analyze the characteristics and laws of online learning behavior. “The mining method model" is a general framework of mining in which three mining objects are aimed at (server log, platform courses database, forum posts), three main mining lines are followed (operation behavior mining flow line, activity behavior mining flow line, verbal behavior mining flow line), and the characteristics of three behaviors are mined (operation behavior mode, activity behavior patterns, verbal behavior characteristics). This model is helpful to find the hidden rules and patterns of online learning from web server system, teaching platform system and learning artifacts library, and it plays an important role in supporting the further study of computer-supported collaborative learning and blended collaborative learning. In this paper, the content of the model is explained in detail, the characteristics of the model are refined and three cases are introduced about how to use the model (namely " mining the path ranking of the visited sites", "analyzing the conversion rate of collaborative cognition processing behavior in BCL", "mining the characteristics of semantic web based on co-word analysis"). The study indicates that the model is feasible and effective.