基于视频点击流数据分析的多时间尺度在线学习行为特征及其调控机制研究  

Research on Multi-Time Scale Online Learning Behavior Characteristics and Their Regulation Mechanisms Based on Video Clickstream Data Analysis

作　　者：刘盼[1] 姜强[1] 张敏[1] 倪静 赵蔚[1] LIU Pan;JIANG Qiang;ZHANG Min;NI Jing;ZHAO Wei(Northeast Normal University,Changchun Jilin 130117;Qingdao Chengyang Experimental High School,Qingdao Shandong 266109)

机构地区：[1]东北师范大学,吉林长春130117 [2]青岛市城阳区实验高级中学,山东青岛266109

出　　处：《现代远距离教育》2024年第6期60-71,共12页Modern Distance Education

基　　金：国家自然科学基金面上项目“网络学习空间中的学习风险预警模型和干预机制研究”(编号:62077012);教育部哲学社会科学研究后期资助项目“高质量发展目标下深度学习的发生机制与促进策略研究”(编号:23JHQ086);吉林省社会科学基金项目“吉林省高等教育数字化转型路径及对策研究”(编号:2023B88);吉林省教育厅社科重点项目“数据驱动的在线学业预警及干预策略研究”(编号:JJKH20241387SK);吉林省高等教育教学改革研究课题“数智时代混合式教学模式下教育技术学专业课程建设与教学实践研究”(编号:2024L5L2IY6001U)。

摘　　要：在数字化转型背景下,数字教育逐渐成为教育改革的核心驱动力。视频作为其重要数字资源,因其直观性和广泛适用性,在重构教育内容、创新教育模式方面发挥着关键作用。然而,如何基于视频点击流数据科学解析在线学习行为规律,从而推动数字教育高质量发展,仍是当前研究亟待解决的关键问题。为此,本研究基于视频点击流数据,结合多时间尺度分析方法与ICAP框架,解析视频点击流行为多时间尺度的动态变化,深入剖析多时间尺度在线学习行为特征差异,包括被动接收、主动操控、建构生成和互动对话的差异性和规律。研究发现,在教学周时间尺度上,行为特征呈现阶段性波动与单双周的显著差异;在教学日时间尺度上,则表现为星期间的周期性变化。在此基础上,从行为生成条件、时间节奏特征和多主体协同维度剖析了在线学习行为在个体认知、课程设计和学习环境维度的影响因素。并从个体自我调整、课程任务分层强化、视频互动引导等方面,提出基于时间节奏优化与行为反馈强化的在线学习行为时间性策略调控机制。In the context of digital transformation,digital education has become a core driving force in educational reform.As a key digital resource,video plays a critical role in reshaping educational content and innovating teaching models due to its intuitiveness and wide applicability.However,how to scientifically analyze online learning behavior patterns based on video data to promote the high-quality development of digital education remains a key issue to be addressed.This study,based on video clickstream data,employs multi-time scale analysis methods and the ICAP framework to analyze the dynamic changes in video clickstream behavior across multiple time scales.It delves into the differences and patterns in multi-time scale online learning behavior characteristics,including passive reception,active manipulation,constructive generation,and interactive dialogue.The findings reveal that,at the weekly teaching scale,behavioral characteristics exhibit periodic fluctuations and significant differences between odd and even weeks,while at the daily teaching scale,they demonstrate cyclical variations across days of the week.On this basis,the study examines the influencing factors of online learning behaviors across individual cognition,course design,and learning environment dimensions from the perspectives of behavioral generation conditions,time rhythm characteristics,and multi-agent collaboration.Furthermore,it proposes a time-based regulation mechanism for online learning behaviors,emphasizing temporal rhythm optimization and feedback enhancement through strategies such as individual self-regulation,hierarchical course task reinforcement,and video interaction guidance.

关 键 词：视频点击流数据 在线学习行为 调控机制 多时间尺度 数字教育 

分 类 号：G43[文化科学—教育学]