Iraqi-E-learning-Emotion-Dataset

This dataset presents an Iraqi E-Learning Emotion Dataset that captures the emotional experiences of students and academic staff toward e-learning in Iraqi universities. The data were collected in 2025 using a voluntary, structured online questionnaire distributed across public and private universities in Iraq. Participants were asked to describe their personal experiences, feelings, and challenges related to e-learning through open-ended free-text responses written in the Iraqi Arabic dialect.
In addition to the textual responses, participants selected one emotion label that best represented their expressed experience from a predefined set of six categories: Joy, Trust, Anticipation, Anger, Anxiety, and Neutral. This design enables the dataset to support supervised emotion classification tasks while preserving the authenticity of self-reported emotional expression.
The dataset consists of 1,182 annotated text entries stored in a UTF-8 encoded CSV file. A standardized preprocessing pipeline was applied to ensure data quality and consistency, including the removal of duplicate entries, URLs, non-text elements, emojis, foreign characters, and normalization of common Iraqi Arabic spelling variations. All preprocessing steps were implemented in Python and are fully documented to support reproducibility.
This dataset addresses a notable gap in Arabic Natural Language Processing resources, as most existing Arabic emotion datasets are derived from social media and do not adequately represent educational contexts. By focusing on e-learning environments in higher education and on a low-resource Arabic dialect, the dataset provides context-specific emotional data that can be reused in research on Arabic emotion analysis, affective computing, educational data mining, learner engagement, and the development of emotion-aware intelligent tutoring systems. The dataset is openly available for research and educational purposes.

本数据集为伊拉克在线学习情感数据集（Iraqi E-Learning Emotion Dataset），用于记录伊拉克高校师生对在线学习的情感体验。数据集于2025年通过自愿参与的结构化在线问卷收集，问卷覆盖伊拉克公立与私立高等院校。参与者需以伊拉克阿拉伯语方言撰写开放式自由文本，描述其与在线学习相关的个人经历、主观感受及面临的挑战。
除上述文本回复外，参与者需从预设的六大情感类别中选择最贴合其表达体验的情感标签：喜悦（Joy）、信任（Trust）、期待（Anticipation）、愤怒（Anger）、焦虑（Anxiety）与中性（Neutral）。该设计使得本数据集可支持监督式情感分类任务，同时保留了自我报告情感表达的真实性。
本数据集共包含1182条标注文本条目，存储于UTF-8编码的CSV文件中。为保障数据质量与一致性，研究团队采用了标准化预处理流程，具体包括删除重复条目、URL链接、非文本元素、表情符号、外来字符，并对常见伊拉克阿拉伯语拼写变体进行归一化处理。所有预处理步骤均通过Python实现，且已完整归档以确保研究可复现性。
当前阿拉伯语自然语言处理（Arabic Natural Language Processing）领域的相关资源存在显著缺口：多数现有阿拉伯语情感数据集均源自社交媒体内容，未能充分覆盖教育场景。本数据集聚焦高等院校在线学习环境与低资源阿拉伯语方言，提供了具有场景针对性的情感数据，可复用于阿拉伯语情感分析、情感计算（affective computing）、教育数据挖掘（educational data mining）、学习者参与度研究以及情感感知型智能教学系统（intelligent tutoring systems）的开发等相关研究。本数据集面向科研与教育用途公开开放。