LMS-FAIR-India

LMS-FAIR-India is a FAIR-compliant synthetic dataset designed to facilitate privacy-preserving research in learning analytics, artificial intelligence, and educational data science. The dataset represents realistic Learning Management System (LMS) interaction patterns derived from a private university academic environment while ensuring that no personally identifiable information (PII) or real user data is exposed. The dataset includes multi-dimensional academic and behavioral records such as student demographics (synthetically generated), course metadata, enrollment mappings, LMS interaction logs (e.g., login events, content access, assignment submissions), assessment outcomes, and derived engagement metrics. These data components are structured in a relational format and accompanied by standardized schemas and metadata to ensure interoperability and reuse. A hybrid synthetic data generation approach was employed, combining statistical distribution modeling and rule-based simulation to preserve the structural, temporal, and behavioral characteristics of real-world LMS systems. The dataset maintains statistical fidelity and supports downstream machine learning and learning analytics applications without compromising user privacy. LMS-FAIR-India adheres to the FAIR principles: Findable: Assigned a persistent identifier (DOI) and indexed metadata Accessible: Freely available via an open repository Interoperable: Provided in standard formats (CSV/JSON) with schema definitions Reusable: Includes comprehensive documentation, data dictionary, and usage guidelines This dataset is intended for researchers, educators, and developers working on: Learning analytics and student behavior modeling AI/ML applications in education Privacy-preserving data mining Educational data benchmarking and reproducibility studies By providing a realistic yet privacy-safe alternative to sensitive educational datasets, LMS-FAIR-India contributes to advancing open science and enabling responsible data sharing in academic ecosystems.LMS-FAIR-India是一款符合FAIR原则的合成数据集，旨在推动学习分析、人工智能及教育数据科学领域的隐私保护研究。该数据集复刻了源自私立高校学术场景的真实学习管理系统（Learning Management System）交互模式，同时确保不会泄露任何个人可识别信息（Personally Identifiable Information，PII）或真实用户数据。 该数据集包含多维度的学术与行为记录，例如人工合成生成的学生人口统计学信息、课程元数据、注册映射关系、LMS交互日志（如登录事件、内容访问、作业提交记录）、评估结果以及衍生的参与度指标。所有数据组件均采用关系型格式进行组织，并配套标准化的模式定义与元数据，以确保数据集的互操作性与可复用性。 研究团队采用了混合式合成数据生成方案，结合统计分布建模与基于规则的模拟技术，以保留真实LMS系统的结构、时序与行为特征。该数据集能够保持统计保真度，可支持下游机器学习与学习分析应用，且不会损害用户隐私。 LMS-FAIR-India严格遵循FAIR原则： - 可发现（Findable）：已分配持久标识符（Digital Object Identifier，DOI）并完成元数据索引 - 可访问（Accessible）：通过开源仓库免费获取 - 可互操作（Interoperable）：采用标准格式（CSV/JSON）提供并附带模式定义 - 可复用（Reusable）：包含完整的文档、数据字典与使用指南 本数据集面向开展以下研究与开发工作的研究人员、教育工作者与开发者： - 学习分析与学生行为建模 - 教育领域的人工智能/机器学习应用 - 隐私保护数据挖掘 - 教育数据基准测试与可重复性研究 通过为敏感教育数据集提供兼具真实性与隐私安全性的替代方案，LMS-FAIR-India有助于推动开放科学发展，并助力学术生态系统中负责任的数据共享实践。