面向实践能力培养的时空大数据分析与挖掘教学改革研究

Teaching Reform of Spatiotemporal Big Data Analysis and Mining for Practice-Oriented Competency Development

摘要: 随着大数据、人工智能与地理空间信息技术的深度融合，时空数据分析能力已成为地理信息类专业人才培养的重要组成部分。针对《时空大数据分析与挖掘》课程中知识内容碎片化、理论与程序实践衔接不足、通用方法与专业场景融合不充分以及评价方式相对单一等问题，本文以实践能力培养为导向，提出“一链三阶、多元评价”的教学改革模式。“一链”以“问题–数据–方法–模型–评价”为完整数据分析流程，重构课程内容；“三阶”按照“基础认知–方法实践–综合应用”组织递进式任务；“多元评价”综合前后测、课堂任务、综合作品和问卷调查评价学习成效。以同一教学班54名学生为对象开展教学实践，结果显示，学生综合成绩由84分提高至90分，课堂任务完成率由85%提高至94%；综合作品中数据处理、模型应用和结果解释三个维度的平均成绩分别为88分、90分和92分，40份有效问卷中各项“同意及非常同意”比例均达到96%以上。多源评价结果表明，该模式有助于强化学生对完整数据分析流程的认识，促进其由程序复现向方法选择、结果解释与自主分析转变，可为地理信息及相关专业数据分析类课程的实践教学改革提供参考。

Abstract: With the deep integration of big data, artificial intelligence, and geospatial information technologies, competence in spatiotemporal data analysis has become an important component of talent development in geographic information and related disciplines. To address fragmented knowledge organization, insufficient connection between theory and programming practice, inadequate integration of general methods with domain scenarios, and relatively single-dimensional assessment in Spatiotemporal Big Data Analysis and Mining, this study proposes a teaching reform model featuring “one task chain, three progressive stages, and multi-source assessment.” The task chain follows the complete workflow of problem definition, data processing, method selection, model construction, and evaluation; the three stages progress from foundational cognition to method-oriented practice and comprehensive application; and learning outcomes are evaluated through pre-/post-tests, classroom tasks, comprehensive student work, and questionnaires. The reform was implemented in one class of 54 students. The overall mean score increased from 84 to 90, while the classroom task completion rate increased from 85% to 94%. For the comprehensive work, the mean scores for data processing, model application, and result interpretation were 88, 90, and 92, respectively; among 40 valid questionnaires, the proportion of students selecting “agree” or “strongly agree” exceeded 96% for all items. Multi-source evaluation results indicate that this model helps strengthen student’ understanding of the complete data analysis workflow, promotes their transition from program reproduction to method selection, result interpretation, and independent analysis, and can provide a reference for practical teaching reform in data analysis courses for geographic information and related majors.