【作 者】：
石琬若， 韩锡斌
【关 键 词 】：
人工智能； 学习分析； 教育中的人工智能； 学习分析与知识国际会议； 文献综述
【栏 目】：
历史与国际比较
【中文摘要】：
2011年，首届学习分析与知识国际会议（简称LAK）召开，标志着学习分析研究领域的确立。2022年，生成式人工智能（AIGC）的爆发，给该领域的研究带来了空前的影响。文章基于最新一届LAK会议的主旨报告及论文，从学习分析的技术系统设计、应用场景、效果评价、伦理风险等方面，系统梳理了AIGC对学习分析领域产生的影响以及今后的发展趋势。研究分析结果表明：基于AIGC的学习分析系统在自动化程度和个性化匹配能力方面得到提升；AIGC具有教师、学习者和助教三种角色的功能属性，围绕数据生成、数据测量与分析、学习反馈等环节拥有多种应用的潜能；从AIGC与传统模型算法的比较、学生学习体验等方面来看，其在学习分析中的应用具有一定的优势，然而尚未有实证证据表明AIGC的使用对学习效果有提升作用。AIGC及其相关技术的发展将为探究人机协同的学习过程带来全新的视角和可能，也将为研究学习过程提供更丰富的数据和创新的方法，同时将促使学习分析价值取向与目标定位的重构，不断推动对复杂人机交互行为方式、人智协同的认知变化与学习规律、AIGC融入的新型教学方式等的理解与探索。
【英文摘要】：
The first Learning Analytics and Knowledge(LAK) conference in 2011 marked the establishment of the research of Learning Analytics. The emergence of Artificial Intelligence Generated Content (AIGC) in 2022 has brought an unprecedented impact on the research in this field. This paper, drawing on keynote addresses and proceedings from the latest LAK conference, systematically reviewed the impact of AIGC on Learning Analytics and its future development trend in terms of technical system design, application scenarios, effectiveness evaluation and ethical risks. The findings indicate that AIGC-based Learning Analytics systems is enhanced in terms of automation and customization matching capabilities. AIGC has the functional attributes of teacher, learner and teaching assistant, and has the potential for various applications around data generation, data measurement and analysis, and learning feedback. From the comparison between AIGC and conventional algorithms, as well as student learning experience, its application in learning analysis has certain advantages. However, there is no empirical evidence yet to suggest that the use of AIGC has an improved effect on learning outcomes. The development of AIGC and its related technologies will bring a new perspective and possibilities for exploring the learning process of human-computer collaboration, and will also provide richer data and innovative methods for studying the learning process. At the same time, it will lead to the reconstruction of the value orientation and goal orientation of learning analysis, and continuously promote the understanding and exploration of complex human-computer interaction behaviors, cognitive changes in human-intelligence collaboration, and learning patterns, as well as new teaching methods integrated with AIGC.