生成式人工智能支持的跨学科主题学习设计模式研究OA

Interdisciplinary thematic learning has attracted considerable attention in educational reform.Although generative artificial intelligence(GenAI)has been preliminarily applied to automated instructional design,its response quality and professional appropriateness in supporting complex design tasks remain inadequate.To address this challenge,this paper constructs a design model of interdisciplinary thematic learning supported by GenAI,in which the"thinking model"is taken as the cognitive support framework of this design model.The proposed model constrains and optimizes the generation logic through four dimensions of domain knowledge,solution procedures,outcome examples,and task criteria.Based on backward design theory and design principles,this paper elaborates the key structure,implementation process,and practical logic of the design model of interdisciplinary thematic learning based on thinking model.On this basis,structured prompts corresponding to each sub-tasks are developed to support instructional design generation under predefined constraints.Empirical results show that GenAI adopting this model achieves outstanding performance in dimensions such as"learning objective design"and"learning assessment design",while room for improvement remains in"learning activity design"and other dimensions that highly rely on contextual judgment.This paper provides a transferable theoretical model and practical implications for human-AI collaborative design of interdisciplinary thematic learning.