MBZUAI/AraSeg-2026-Shared-Task-PA

AraSeg是第一个全面的阿拉伯语句子分割基准语料库，旨在支持现代标准阿拉伯语（MSA）的句子分割研究，特别是在标点符号不一致、缺失或嘈杂的环境中。该数据集包含从不同来源和体裁收集的手动标注文档，支持跨不同写作风格和领域的鲁棒评估。AraSeg-PA是语料库的段落感知（PA）变体，其中文档包括段落边界和标点符号。任务定义为二元标记分类任务，预测每个标记后是否跟随句子边界。AraSeg is the first comprehensive benchmark for Arabic sentence segmentation. The corpus is designed to support research on sentence segmentation in Modern Standard Arabic (MSA), particularly in settings where punctuation is inconsistent, missing, or noisy. AraSeg contains manually annotated documents collected from diverse sources and genres, enabling robust evaluation across different writing styles and domains. AraSeg-PA is the Paragraph-Aware (PA) variant of the corpus where documents include paragraph boundaries and punctuation marks. The task is formulated as a binary token classification task to predict sentence boundaries.