metadata

license: mit
language:
  - ar
  - sq
  - am
  - az
  - bn
  - bs
  - bg
  - my
  - zh
  - da
  - nl
  - en
  - tl
  - fr
  - ff
  - fa
  - de
  - gu
  - ha
  - hi
  - id
  - it
  - ja
  - jv
  - kk
  - km
  - ko
  - ku
  - ky
  - ms
  - ml
  - ps
  - pl
  - pt
  - pa
  - ru
  - sd
  - si
  - so
  - es
  - sw
  - sv
  - tg
  - ta
  - tt
  - te
  - th
  - tr
  - ur
  - ug
  - uz
  - yo
  - 'no'
  - vi
task_categories:
  - translation

📘 Qur’an Multilingual Parallel Dataset (`quran_multilingual_parallel`)

This dataset presents a clean, structurally-aligned multilingual parallel corpus of the Qur’anic text. It is intended for linguistic, computational, and cross-lingual AI applications — not only for religious interpretation.

It contains over 6,200 verse-level alignments in 54 human languages, formatted in a machine-friendly .csv structure with language-specific translation fields.

🧠 Dataset Highlights

📖 6,236 ayahs (verses)
📚 114 surahs (chapters)
🌍 Translations in 53 languages
🕌 Arabic source text included
🔢 Fully aligned, row-per-verse CSV
⚠️ Missing translations are transparently documented

📁 Files Included

File	Description
`quran_multilingual_parallel.csv`	Complete parallel corpus: 6,236 rows × 57 columns
`README.md`	This documentation

🧾 Column Format

Each row corresponds to a verse (ayah). Columns include:

Column	Description
`surah_number`	Integer: Surah (chapter) number (1–114)
`surah`	Slugified Surah name (e.g., `al-baqarah`)
`ayah`	Integer: Ayah (verse) number
`arabic`	Original Arabic text of the verse
`{language}`	One column per language, each containing the translation for that ayah

🌍 Languages Included

This dataset includes translations in the following 54 languages:

1.	Arabic         (original)
2.	Albanian
3.	Amharic
4.	Azerbaijani
5.	Bengali
6.	Bosnian
7.	Bulgarian
8.	Burmese
9.	Chinese
10.	Danish
11.	Dutch
12.	English
13.	Filipino
14.	French
15.	Fulah
16.	Persian
17.	German
18.	Gujarati
19.	Hausa
20.	Hindi
21.	Indonesian
22.	Italian
23.	Japanese
24.	Javanese
25.	Kazakh
26.	Khmer
27.	Korean
28.	Kurdish
29.	Kyrgyz
30.	Malay
31.	Malayalam
32.	Norwegian
33.	Pashto
34.	Polish
35.	Portuguese
36.	Punjabi
37.	Russian
38.	Sindhi
39.	Sinhalese
40.	Somali
41.	Spanish
42.	Swahili
43.	Swedish
44.	Tajik
45.	Tamil
46.	Tatar
47.	Telugu
48.	Thai
49.	Turkish
50.	Urdu
51.	Uyghur
52.	Uzbek
53.	Vietnamese
54.	Yoruba

⚠️ Missing Translation Entries

While all 6,236 verses are structurally included, 79 translations were unavailable from the source site (al-quran.cc) at the time of scraping.

These translations are missing in the following (language, surah, ayah) combinations:

📋 Click to expand full list of missing translations (79)

azerbaijani — al-asr:1
azerbaijani — al-asr:2
azerbaijani — al-asr:3
filipino — al-haaqqa:52
german — al-baqara:220
german — al-isra:65
german — an-naba:3
german — ar-rum:5
german — az-zamar:40
jawa — al-furqan:37
kyrgyz — al-mulk:16
portuguese — aal-e-imran:50
portuguese — aal-e-imran:51
portuguese — an-naml:31
portuguese — an-naziat:19
portuguese — as-saaffat:53
portuguese — ash-shuara:17
portuguese — nooh:11
portuguese — nooh:12
portuguese — nooh:13
portuguese — nooh:14
portuguese — nooh:15
portuguese — nooh:16
portuguese — nooh:17
portuguese — nooh:18
portuguese — nooh:19
portuguese — nooh:20
sindhi — ibrahim:1
sindhi — ibrahim:2
sindhi — ibrahim:3
sindhi — ibrahim:4
sindhi — ibrahim:5
sindhi — ibrahim:6
sindhi — ibrahim:7
sindhi — ibrahim:8
sindhi — ibrahim:9
sindhi — ibrahim:10
sindhi — ibrahim:11
sindhi — ibrahim:12
sindhi — ibrahim:13
sindhi — ibrahim:14
sindhi — ibrahim:15
sindhi — ibrahim:16
sindhi — ibrahim:17
sindhi — ibrahim:18
sindhi — ibrahim:19
sindhi — ibrahim:20
sindhi — ibrahim:21
sindhi — ibrahim:22
sindhi — ibrahim:23
sindhi — ibrahim:24
sindhi — ibrahim:25
sindhi — ibrahim:26
sindhi — ibrahim:27
sindhi — ibrahim:28
sindhi — ibrahim:29
sindhi — ibrahim:30
sindhi — ibrahim:31
sindhi — ibrahim:32
sindhi — ibrahim:33
sindhi — ibrahim:34
sindhi — ibrahim:35
sindhi — ibrahim:36
sindhi — ibrahim:37
sindhi — ibrahim:38
sindhi — ibrahim:39
sindhi — ibrahim:40
sindhi — ibrahim:41
sindhi — ibrahim:42
sindhi — ibrahim:43
sindhi — ibrahim:44
sindhi — ibrahim:45
sindhi — ibrahim:46
sindhi — ibrahim:47
sindhi — ibrahim:48
sindhi — ibrahim:49
sindhi — ibrahim:50
sindhi — ibrahim:51
sindhi — ibrahim:52

These empty fields are left intentionally blank in the dataset. All ayahs are present in Arabic, and structural alignment is maintained.

🔍 Data Source and Extraction

Source: https://www.al-quran.cc/quran-translation/
Extraction: Python-based HTML parsing with dynamic AJAX handling
Structural validation: 114 surahs × 6,236 ayahs cross-checked
Missing translations logged during post-validation

📜 Linguistic and Literary Significance of the Qur’an

🧬 A Living Corpus Preserved in Speech

The Qur’an is the only major historical text that has been memorized word-for-word by millions of people across generations, regions, and languages — regardless of whether they spoke Arabic natively.

Over 1,400 years old, the Arabic text of the Qur’an remains unchanged, recited daily, and actively memorized, verse by verse.
It is preserved not only in manuscripts but in oral transmission, making it one of the most reliably reconstructed texts in human linguistic history.

🌍 Translated into Languages That Weren’t Yet Standardized

Several languages in this dataset — such as today’s form of Burmese (Myanmar), Swahili, Standard Indonesian, and even Modern English — had not yet developed or reached their current standardized written form at the time the Qur’an was first revealed over 1,400 years ago.

Yet today, these communities:

Study and translate the Qur’an with deep linguistic care
Retain the verse structure in all translations
Participate in a global multilingual alignment with the same source text

This makes the Qur’an a linguistic anchor across centuries, preserved in meaning and form across dozens of linguistic systems.

✍️ Neither Prose Nor Poetry — A Unique Register

From a literary standpoint, the Qur’an’s structure is:

Not classical prose
Not traditional poetry
But a distinct rhythmic and rhetorical form known for:
- Internal rhyme and parallelism
- Recurring motifs
- High semantic density and emotional resonance

This genre continues to challenge both literary analysis and computational modeling.

🤖 Relevance to AI and Linguistics

The Qur’an is an ideal corpus for AI-based multilingual NLP:

🔁 Fully aligned across 53 languages
🧩 Rigid source structure (verse-level)
🧠 Oral memory transmission modeling
🌐 Cross-lingual semantic drift analysis
🛠️ Faithful translation alignment tasks

It provides a fixed point of semantic comparison for understanding how languages represent meaning — historically and today.

🪪 License

Text: Public Domain where applicable (verify by language)
Code/scripts: MIT License

📌 Citation

@misc{quran_multilingual_parallel_2025,
  title  = {Qur’an Multilingual Parallel Dataset},
  author = {freococo},
  year   = {2025},
  url    = {https://huggingface.co/datasets/freococo/quran_multilingual_parallel}
}

📘 Qur’an Multilingual Parallel Dataset (quran_multilingual_parallel)