Al-Mawrid-US/almawrid-books

Al-Mawrid Books Corpus

Full-text content from Al-Mawrid's published book library (English, Urdu, and Arabic). Each row is one complete book — all index chunks merged into a single record.

Source: Al-Mawrid PDF library. One row per book title.

Data format

Each corpus is exported in two merged forms (one row per complete document):

File	Purpose
data/books.json	Pretty-printed JSON array — sprint export with human-readable field names
data/books.csv	UTF-8 CSV with BOM — comma-separated, quoted fields, for Excel and Hugging Face

Intermediate chunk files (books_chunks.jsonl) store exact Azure index documents (PascalCase field names, all retrievable fields, no embeddings) and are not uploaded. Multi-chunk sources (books, YouTube) are merged into a single row before export. Index-only fields (document_id, content_type, chunk_index, embeddings) are excluded from the public release.

CSV files use UTF-8 with BOM, comma delimiter, and quoted fields so Arabic, Urdu, and English display correctly in Excel and the Hugging Face dataset viewer.

Field descriptions

Column	Type	Description
title	string	Title of the item (forum thread title, book title, or video title)
text	string	Full merged content (Q&A: answer; book: full text; YouTube: full transcript)
source_url	string	Link back to the original source (forum URL, PDF, or YouTube video)
language	string	ISO-style language code (en, ur, ar) or unknown

Python usage example

from datasets import load_dataset

ds = load_dataset("Al-Mawrid-US/almawrid-books", split="train")

# Browse the first item
print(ds[0]["title"])
print(ds[0]["text"][:500])

# Filter by language
english = ds.filter(lambda x: x["language"] == "en")
print(f"English records: {len(english)}")

Licence & permitted uses

This dataset is released under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence.

You are free to share and adapt the data for any purpose, including commercial use, provided that you give appropriate credit to Al-Mawrid, provide a link to the licence, and indicate if changes were made.

Citation

If you use this dataset in your work, please cite:

@dataset{almawrid_books_2026,
  author       = {Al-Mawrid},
  title        = {Al-Mawrid Books Corpus},
  year         = {2026},
  publisher    = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/Al-Mawrid-US/almawrid-books}},
  license      = {CC BY 4.0}
}

Provenance

Exported from the Al-Mawrid Azure AI Search knowledge index. Chunk-level index fields and embeddings are excluded from this public release.