metadata

language:
  - ar
multilinguality:
  - monolingual
task_categories:
  - summarization
task_ids:
  - news-articles-summarization
  - language-modeling
pretty_name: Arabic News
dataset_info:
  features:
    - name: Body
      dtype: string
    - name: Title
      dtype: string
  splits:
    - name: train
      num_bytes: 1479976249
      num_examples: 588881
  download_size: 745147868
  dataset_size: 1479976249
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags: []

Arabic News Dataset

Dataset Summary

The Arabic News Dataset contains a collection of news articles in Arabic, featuring titles and corresponding bodies of text. It is designed for natural language processing tasks such as text classification, summarization, and sentiment analysis.

Dataset Details

Languages: Arabic
Number of Samples: [6.15k]
Features:
- title: The title of the news article.
- body: The full text of the news article.

Usage

This dataset can be used for various NLP tasks, including:

Text classification (e.g., categorizing news articles)
Summarization (e.g., generating summaries of news articles)
Sentiment analysis (e.g., determining the sentiment of news articles)

Dataset Splits

Train: [Number of training samples]
Validation: [Number of validation samples]

Data Collection

This dataset was collected from various Arabic news sources, ensuring a diverse range of topics and styles.