KFUPM-JRCAI/arabic-generated-abstracts

该数据集是一个阿拉伯语机器生成文本数据集，包含了多种生成方法产生的文本，用于支持机器生成文本检测系统的研究、风格分析以及跨模型泛化研究。数据集分为根据标题、根据标题和内容以及通过润色三种生成方法的子集，每个样本包括原始人类摘要和四种不同模型生成的摘要。

This dataset is an Arabic machine-generated text dataset containing text produced by various generation methods, supporting research in machine-generated text detection systems, stylometric analysis, and cross-model generalization studies. The dataset is divided into subsets based on title-only, title and content, and polishing generation methods, with each sample including the original human-written abstract and abstracts generated by four different models.