- 期刊
聚焦目标适性达成：自适应学习领域模型的国际研究进展
Focusing on Goal Adaptive Achievement: International Progress in Domain Model Research for Adaptive Learning
摘要
自适应学习是人工智能教育应用的重要形式。由领域模型、学习者模型、导学模型和人机交互模型作为核心组件构成自适应学习环境，而领域模型指向适应性的学习目标达成，是自适应学习的基础，近年来成为研究的热点问题。鉴于此，聚焦近五年领域模型的国际进展，分析发现，领域模型呈现从良构领域到混合领域的发展趋势，主要包括目标领域表示和领域导学策略两大研究方向。已有研究存在以结构化知识点、认知过程和本体属性三种表示方式，具体从基于学习支架、基于启发对话和基于学习者模型三方面设计导学策略，利用领域知识抽取，领域内容生成和领域内容优化等技术助力领域模型的构建。然而，当前研究还面临领域碎片化、粒度难把控、未适应全人培养和教育认同缺失等现实挑战，未来需在开放共享、满足差异、走向异质化和策略多样化等方面进一步开展研究。
並列摘要
Adaptive learning is an important form of AI education applications. Domain model, student model, tutor model and human-computer interaction model constitute the adaptive learning environment as the core components. The domain model points to the achievement of adaptive learning goals, which is the basis of adaptive learning and has become a hot research topic in recent years. In view of this, this study focuses on the international progress of domain models in the past five years. The analysis finds that the domain model shows a development trend from well-structured domain to mixed-structured domain, mainly including two research directions: target domain representation and domain tutoring strategy. In the existing research, there are three ways to represent objective domain: structured knowledge points, cognitive processes and ontology attributes. Learning guidance strategies are designed from three aspects: based on learning scaffolds, based on heuristic dialogues, and based on learner model. Technologies such as domain content extraction, domain content generation and domain content optimization were used to construct domain models. However, the current research still faces practical challenges such as fragmentation of the field, difficult to control granularity, failure to adapt to the cultivation of the whole person, and lack of educational identity. Further research will be conducted on open sharing, satisfying differences, moving towards heterogeneity and strategy diversification.