README.md

metadata

dataset_info:
  features:
    - name: non_vocalized
      dtype: string
    - name: vocalized
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 3609941776
      num_examples: 1463790
    - name: valid
      num_bytes: 74699622
      num_examples: 30181
    - name: test
      num_bytes: 37176837
      num_examples: 15091
  download_size: 3516498736
  dataset_size: 3721818235
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: valid
        path: data/valid-*
      - split: test
        path: data/test-*
license: mit
language:
  - ar
pretty_name: Arabic Tashkeel Dataset

Arabic Tashkeel Dataset

This is a fairly large dataset gathered from five main sources:

• tashkeela (1.79GB - 45.05%): The entire Tashkeela dataset, repurposed in sentences. Some rows were omitted as they contain low diacritic (tashkeel characters) rate.
• shamela (1.67GB - 42.10%): Random pages from over 2,000 books on the Shamela Library. Pages were selected using the below function (high diacritics rate)
• wikipedia (269.94MB - 6.64%): A collection of Wikipedia articles. Diacritics were added using OpenAI's GPT-4o mini model. At the time of writing this, many other LLMs were tried (such as GPT-4o, Claude 3 Haiku, Claude 3.5 Sonnet, Llama 3.1 70b, among others), and this one (surprisingly) scored the highest in a subset of the tashkeela dataset.
• ashaar (117.86MB - 2.90%): APCD, APCDv2, Ashaar_diacritized, Ashaar_meter merged. Most rows from these datasets were excluded, and only those with sufficient diacritics were retained.
• quran-riwayat (71.73MB - 1.77%): Six different riwayat of Quran.
• hadith (62.69MB - 1.54%): Leeds University and King Saud University (LK) Hadith Corpus.

To filter out samples that contain partial or no tashkeel, we only retain sentences where diacritic characters make up 70% or more of the Arabic characters, using this function:

import pyarabic.araby as araby

# a function that determines whether a text contains Arabic diacritics
def has_diacritics(text):
    tashkeel_chars = set(araby.TASHKEEL)
    arabic_chars = set("ابتثجحخدذرزسشصضطظعغفقكلمنهويىءآأؤإئ")
    # if tashkeel characters's count is greater than 70% of the Arabic characters' count, then the text has diacritics
    return sum(1 for c in text if c in tashkeel_chars) >= 0.7 * sum(1 for c in text if c in arabic_chars)

We make use of the pyarabic library, make sure to install it:

$ pip install pyarabic

Main Uses

This dataset can be used to train models to automatically add diacritics (perform tashkeel) to Arabic text.

Limitations

Over 90% of the dataset consists primarily of religious texts in Classical Arabic. As a result, models trained on this data are well-suited for vocalizing such texts but may struggle with Modern Standard Arabic. Wikipedia articles were added to help ease this issue, though they may not entirely resolve it.