生成式人工智能赋能商科数据课程群教学改革

Generative Artificial Intelligence Empowering the Teaching Reform of a Business Data Curriculum Cluster

摘要: 针对商科数据类课程内容割裂、技术训练与商务决策脱节、学生基础差异及生成式人工智能使用过程难以评价等问题，文章以“大数据处理技术及商务应用”和“机器学习与商务智能”为研究对象，基于建构性对齐与项目化学习理念，构建课程群项目化教学模式。模式将AI认知支架界定为随任务提供、可核验且逐步撤除的支持，按工具类型明确使用边界，并以客户流失项目区分基础验证、综合应用、开放探究和反思答辩四类任务。将AI输出核验列为学习目标，设置“AI勘误”练习，运用单元测试、交叉验证和来源核查形成证据。以贯通项目连接数据治理、分析建模与决策表达，建立覆盖学习过程、项目成果、个体掌握、核验能力和规范素养的评价体系，为商科数据课程群改革提供可实施、可追溯、可检验的设计方案。

Abstract: To address fragmented course content, the disconnect between technical training and business decision-making, differences in students’ prior knowledge, and difficulties in evaluating the use of generative artificial intelligence (AI), this study develops a project-based course-cluster teaching model for “Big Data Processing Technologies and Business Applications” and “Machine Learning and Business Intelligence”. Drawing on constructive alignment and project-based learning, the model defines AI-based cognitive scaffolding as task-responsive, verifiable support that is gradually withdrawn, and specifies usage boundaries for different types of AI tools. Using a customer churn project, it distinguishes four types of tasks: foundational verification, integrated application, open-ended inquiry, and reflective defense. The verification of AI-generated outputs is established as an explicit learning objective. An “AI error-detection and correction” exercise is introduced, with unit testing, cross-validation, and source verification employed to generate supporting evidence. Through a continuous project linking data governance, analytical modeling, and decision communication, the model establishes an assessment system covering the learning process, project outcomes, individual mastery, verification competence, and ethical and responsible AI use. It thereby provides an implementable, traceable, and testable design for reforming business data course clusters.