ES-KT-24
收藏资源简介:
ES-KT-24是由Enuma, Inc.和韩国大学联合创建的多模态知识追踪基准数据集,旨在为智能教育系统提供丰富的游戏化学习环境数据。该数据集包含教育游戏视频、合成问题文本和详细的游戏日志,涵盖数学、英语、印尼语和马来语等多个学科。数据集通过大型语言模型生成合成文本,包含28个知识概念和182个问题,涉及15,032名用户和7,782,928次互动。数据集的创建过程包括游戏录制、文本生成和数据预处理,旨在通过多模态数据分析学生学习模式,推动知识追踪和学习分析领域的研究。
ES-KT-24 is a multimodal knowledge tracing benchmark dataset jointly developed by Enuma, Inc. and Korea University, designed to provide rich gamified learning environment data for intelligent education systems. This dataset comprises educational game videos, synthesized question texts and detailed game logs, covering multiple disciplines including mathematics, English, Indonesian and Malay. The synthesized texts, generated via large language models, contain 28 knowledge concepts and 182 questions, involving 15,032 users and 7,782,928 interaction records. The dataset's creation workflow includes game recording, text generation and data preprocessing, aiming to analyze students' learning patterns through multimodal data and advance research in the fields of knowledge tracing and learning analytics.




