- 期刊
基于情境—认知—技术三重视角的智能测评诊断框架、实践和应用
A Diagnostic Framework, Practice Path and Application for Intelligent Assessment Based on Context-Cognition-Technology Perspective
摘要
智能测评诊断建模已成为洞察学习者认知发展的新范式，对教育数字化转型具有重要的理论价值和实践意义。面对当前的教育情境复杂性和认知过程易变性挑战，构建基于情境、认知和技术三重視角的测评诊断框架，有助于挖掘从教育情境现象到内在认知的发展规律，揭示学习者认知随时间推移的演变机理。在情境视角下，构建“知识－交互－行为－时序”四元组，实现教育测评的全要素、全过程性挖掘；在认知视角下，针对四元要素开展速度、记忆和综合能力等认知规律建模，探索四元教育要素间的关联；在技术视角下，借助深度学习技术，探索基于多维交互数据、时序过程信息和综合能力的测评诊断实践路径，推断学习者的认知状态。通过可视化诊断分析结果，开展资源推荐、路径规划和教学干预等教学应用，促进数据驱动的大規模精准教学和个性化自主学习的实施。
並列摘要
The implementation of intelligent assessment diagnostic modeling has emerged as a novel paradigm that offers crucial insights into the cognitive growth of learners. This paradigm bears immense theoretical and practical significance in facilitating the digital transformation of education. Faced with the challenges of the complexity of educational situations and the variability of cognitive processes, constructing an intelligent assessment and diagnosis framework that encompasses context, cognition, technology perspective will help to explore the development law from educational situation phenomena to internal cognition, and reveal the evolution mechanism of learners' cognition over time. In the contextual perspective, we can establish the four-tuple group of "knowledge-interaction-behavior-time-sequence" to facilitate the mining of all elements and the entire process of educational assessment. In the cognitive perspective, we can model cognitive laws such as speed, memory, and comprehensive abilities for the four elements, and explore the relationship between education elements. In the technical perspective, we can utilize deep learning technology to explore the practice path of assessment and diagnosis based on interactive data standards, time series process information, and comprehensive abilities, so as to infer the cognitive state of learners. Through visual diagnosis and analysis results, teaching applications such as resource recommendation, path planning and teaching intervention are carried out to promote the implementation of data-driven large-scale precision teaching and personalized autonomous learning.