面向GeoAI的《时空大数据分析》课程教学改革研究与实践

Exploration and Practice on Teaching Reform of the “Spatio-Temporal Big Data Analysis” Course Oriented to GeoAI

摘要: 地理空间人工智能(GeoAI)正在推动地理信息科学人才培养从传统GIS技能训练转向数据、算法、平台与地理问题综合解决能力培养。《时空大数据分析》课程处于地理信息科学专业基础课程与GeoAI前沿应用之间的关键衔接位置，但在本科教学中仍面临知识跨度大、教学资源适配性不足、实践环境门槛高和案例专业指向不够突出等问题。本文基于辽宁师范大学地理科学学院本科教学改革项目，围绕“AI + 地理”综合素养培养目标，构建“问题牵引、内容重构、资源支撑、平台实践、评价改进”的课程改革框架。改革方案以大单元教学重组课程内容，建设线上教学资源库、地理时空数据库和教学案例库，并依托JupyterHub平台形成“教、学、练、战”一体化实践环境。研究认为，面向GeoAI的课程改革应避免简单移植人工智能算法课程，而应从地理问题、时空数据、智能方法和实践平台协同发力，推动学生在真实地理场景中形成数据处理、模型应用、结果解释和创新实践能力。该改革路径可为地理信息科学专业复合型应用人才培养提供课程建设参考。

Abstract: Geospatial artificial intelligence (GeoAI) is promoting the transformation of talent cultivation in geographic information science from traditional GIS skills training toward the integrated development of data, algorithms, platforms, and geographic problem-solving competence. The course “Spatio-temporal Big Data Analysis” plays a key bridging role between foundational courses in geographic information science and frontier GeoAI applications. However, undergraduate teaching still faces problems such as broad knowledge coverage, insufficient adaptation of teaching resources, high barriers to practice environments, and inadequate professional orientation of cases. Based on the undergraduate teaching reform project of the School of Geography, Liaoning Normal University, this paper focuses on the cultivation goal of “AI + geography” comprehensive literacy and constructs a curriculum reform framework featuring problem orientation, content reconstruction, resource support, platform-based practice, and evaluation improvement. The reform reorganizes course content through large-unit teaching, builds an online teaching resource repository, a geographic spatio-temporal database, and a teaching case library, and relies on JupyterHub to establish an integrated practice environment for teaching, learning, training, and project-based application. The study argues that GeoAI-oriented curriculum reform should avoid simply transplanting artificial intelligence algorithm courses. Instead, it should coordinate geographic problems, spatio-temporal data, intelligent methods, and practice platforms to help students develop abilities in data processing, model application, result interpretation, and innovative practice in real geographic scenarios. This reform pathway can provide reference of curriculum development for cultivating interdisciplinary applied talents in geographic information science.