README.md

metadata

configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
dataset_info:
  features:
    - name: GenId
      dtype: int64
    - name: SubId
      dtype: int64
    - name: DatasetName
      dtype: string
    - name: DatasetLink
      dtype: string
    - name: Text
      dtype: string
    - name: MetaData
      struct:
        - name: AboutAuthor
          dtype: string
        - name: AboutBook
          dtype: string
        - name: Author
          dtype: string
        - name: AuthorName
          dtype: string
        - name: BookLink
          dtype: string
        - name: BookName
          dtype: string
        - name: ChapterLink
          dtype: string
        - name: ChapterName
          dtype: string
        - name: Tags
          dtype: float64
        - name: __index_level_0__
          dtype: float64
        - name: created_date
          dtype: string
        - name: deleted
          dtype: bool
        - name: detoxify
          dtype: 'null'
        - name: emojis
          struct:
            - name: count
              sequence: int32
            - name: name
              sequence: string
        - name: id
          dtype: string
        - name: labels
          struct:
            - name: count
              sequence: int32
            - name: name
              sequence: string
            - name: value
              sequence: float64
        - name: lang
          dtype: string
        - name: message_id
          dtype: string
        - name: message_tree_id
          dtype: string
        - name: model_name
          dtype: 'null'
        - name: parent_id
          dtype: string
        - name: rank
          dtype: float64
        - name: review_count
          dtype: float64
        - name: review_result
          dtype: bool
        - name: role
          dtype: string
        - name: synthetic
          dtype: bool
        - name: title
          dtype: string
        - name: tree_state
          dtype: string
        - name: url
          dtype: string
        - name: user_id
          dtype: string
    - name: ConcatenatedText
      dtype: int64
    - name: __index_level_0__
      dtype: int64
  splits:
    - name: train
      num_bytes: 1955979492
      num_examples: 127898
  download_size: 774823370
  dataset_size: 1955979492
task_categories:
  - conversational
  - text-generation
  - text2text-generation
  - translation
  - summarization
language:
  - ar
pretty_name: MAD

Dataset Card for "Mixed-Arabic-Dataset"

Mixed Arabic Datasets (MAD)

The Mixed Arabic Datasets (MAD) project provides a comprehensive collection of diverse Arabic-language datasets, sourced from various repositories, platforms, and domains. These datasets cover a wide range of text types, including books, articles, Wikipedia content, stories, and more.

MAD Repo vs. MAD Main

MAD Repo

• Versatility: In the MAD Repository (MAD Repo), datasets are made available in their original, native form. Researchers and practitioners can selectively download specific datasets that align with their specific interests or requirements.
• Independent Access: Each dataset is self-contained, enabling users to work with individual datasets independently, allowing for focused analyses and experiments.

MAD Main or simply MAD

• Unified Dataframe: MAD Main represents a harmonized and unified dataframe, incorporating all datasets from the MAD Repository. It provides a seamless and consolidated view of the entire MAD collection, making it convenient for comprehensive analyses and applications.
• Holistic Perspective: Researchers can access a broad spectrum of Arabic-language content within a single dataframe, promoting holistic exploration and insights across diverse text sources.

Why MAD Main?

• Efficiency: Working with MAD Main streamlines the data acquisition process by consolidating multiple datasets into one structured dataframe. This is particularly beneficial for large-scale projects or studies requiring diverse data sources.
• Interoperability: With MAD Main, the datasets are integrated into a standardized format, enhancing interoperability and compatibility with a wide range of data processing and analysis tools.
• Meta-Analysis: Researchers can conduct comprehensive analyses, such as cross-domain studies, trend analyses, or comparative studies, by leveraging the combined richness of all MAD datasets.

Getting Started

• To access individual datasets in their original form, refer to the MAD Repository (Link to MAD Repo).
• For a unified view of all datasets, conveniently organized in a dataframe, you are here in the right place.

from datasets import load_dataset

dataset = load_dataset("M-A-D/Mixed-Arabic-Dataset-Main")

Join Us on Discord

For discussions, contributions, and community interactions, join us on Discord!

How to Contribute

Want to contribute to the Mixed Arabic Datasets project? Follow our comprehensive guide on Google Colab for step-by-step instructions: Contribution Guide.

Note: If you'd like to test a contribution before submitting it, feel free to do so on the MAD Test Dataset.

Citation

@dataset{ 
title = {Mixed Arabic Datasets (MAD)},
author = {MAD Community},
howpublished = {Dataset},
url = {https://huggingface.co/datasets/M-A-D/Mixed-Arabic-Datasets-Repo},
year = {2023},
}