生成式人工智能能够根据文本提示或音乐素材生成旋律、配器及音频片段,降低了音乐创作的技术门槛,也使仅以最终作品为中心的评价面临新的解释问题:作品完成度并不必然反映学习者的创作思考与审美判断。本文采用针对性叙述检索与理论建构方法,围绕高校音乐教育专业学生的AI辅助音乐创作,辨析作品生成能力、创作参与能力与创作主体性三个相关概念,并从创作意图、审美判断、主动修改、反思表达四个方面提出过程性评价分析框架。在此基础上,本文讨论生成式人工智能介入创作过程后形成的评价张力,并提出意图前置、关键过程证据收集、多元反馈与教师引导等建议。本文提出的框架仅是初步理论分析工具,尚未经过课堂实证检验,其适用性、维度关系及可操作性仍有待后续研究考察。
生成式人工智能;音乐教育;音乐创作;创作主体性;过程性评价
Generative artificial intelligence can produce melodies, arrangements, and audio works from textual prompts or musical materials. While lowering the technical threshold of music creation, it also challenges outcome-oriented assessment: the completeness of a generated work does not necessarily represent learners’ creative thinking or aesthetic judgment. Based on a literature review and theoretical construction, this paper examines AI-assisted music creation among university students majoring in music education. It distinguishes work-generation ability, creative participation, and creative subjectivity, and proposes a process-oriented analytical framework comprising four dimensions: creative intention, aesthetic judgment, active modification, and reflective articulation. The paper further discusses the assessment tensions arising from generative AI and proposes instructional strategies involving prior articulation of intentions, collection of process evidence, multiple forms of feedback, and teacher guidance. The framework is intended as a preliminary theoretical tool for analyzing learner agency in AI-assisted music creation. As it has not yet been empirically tested in classroom settings, its applicability and the relationships among its dimensions require further investigation.
Generative Artificial Intelligence; Music Education; Music Creation; Creative Subjectivity; Process Evaluation
黄世英,易秀华. 生成式人工智能辅助音乐创作的过程性评价框架研究——基于创作主体性四维分析[J]. 中国现代教育学报. 2026, 2 (6): 34-40. DOI: 10.70693/202610276075.
黄世英, 易秀华. (2026). 生成式人工智能辅助音乐创作的过程性评价框架研究——基于创作主体性四维分析. 中国现代教育学报, 2 (6), 34-40. https://doi.org/10.70693/202610276075