drelhaj/EASC

EASC: The Essex Arabic Summaries Corpus Mo El-Haj, Udo Kruschwitz, Chris FoxUniversity of Essex, UK This repository hosts EASC — the Essex Arabic Summaries Corpus — a collection of 153 Arabic source documents and 765 human-generated extractive summaries, created using Amazon Mechanical Turk. EASC is one of the earliest publicly available datasets for Arabic single-document summarisation and remains widely used in research on Arabic NLP, extractive summarisation, sentence… See the full description on the dataset page: https://huggingface.co/datasets/drelhaj/EASC.