README.md

---
language:
- ar
- en
license: cc-by-4.0
task_categories:
- question-answering
pretty_name: Quran Question Answer with Context
dataset_info:
  features:
  - name: q_id
    dtype: int64
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: q_word
    dtype: string
  - name: q_topic
    dtype: string
  - name: fine_class
    dtype: string
  - name: class
    dtype: string
  - name: ontology_concept
    dtype: string
  - name: ontology_concept2
    dtype: string
  - name: source
    dtype: string
  - name: q_src_id
    dtype: int64
  - name: quetion_type
    dtype: string
  - name: chapter_name
    dtype: string
  - name: chapter_no
    dtype: int64
  - name: verse
    dtype: string
  - name: answer_en
    dtype: string
  - name: class_en
    dtype: string
  - name: fine_class_en
    dtype: string
  - name: ontology_concept2_en
    dtype: string
  - name: ontology_concept_en
    dtype: string
  - name: q_topic_en
    dtype: string
  - name: q_word_en
    dtype: string
  - name: question_en
    dtype: string
  - name: chapter_name_en
    dtype: string
  - name: verse_list
    sequence: int64
  - name: context
    dtype: string
  - name: context_data
    dtype: string
  - name: context_missing_verses
    dtype: string
  splits:
  - name: train
    num_bytes: 3534771
    num_examples: 1224
  download_size: 1858762
  dataset_size: 3534771
tags:
- islam
- quran
- arabic
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
# Dataset Card for "quran-question-answer-context"

## Dataset Summary

Translated the original dataset from Arabic to English and added the Surah ayahs to the `context` column.

## Usage

```python
from datasets import load_dataset

dataset = load_dataset("nazimali/quran-question-answer-context")
```

```python
DatasetDict({
    train: Dataset({
        features: ['q_id', 'question', 'answer', 'q_word', 'q_topic', 'fine_class', 'class', 'ontology_concept', 'ontology_concept2', 'source', 'q_src_id', 'quetion_type', 'chapter_name', 'chapter_no', 'verse', 'answer_en', 'class_en', 'fine_class_en', 'ontology_concept2_en', 'ontology_concept_en', 'q_topic_en', 'q_word_en', 'question_en', 'chapter_name_en', 'verse_list', 'context', 'context_data', 'context_missing_verses'],
        num_rows: 1224
    })
})
```

## Translation Info

1. Translated the Arabic questions/concept columns to English with [Helsinki-NLP/opus-mt-ar-en](https://huggingface.co/Helsinki-NLP/opus-mt-ar-en)
2. Used `en-yusufali` translations for ayas [M-AI-C/quran-en-tafssirs](https://huggingface.co/datasets/M-AI-C/quran-en-tafssirs)
3. Renamed Surahs with [kheder/quran](https://huggingface.co/datasets/kheder/quran)
4. Added the ayahs that helped answer the questions
  - Split the `ayah` columns string into a list of integers
  - Concactenated the Surah:Ayah pairs into a sentence to the `context` column

Columns with the suffix `_en` contain the translations of the original columns.

## TODO
The `context` column has some `null` values that needs to be investigated and fixed

## Initial Data Collection

The original dataset is from **[Annotated Corpus of Arabic Al-Quran Question and Answer](https://archive.researchdata.leeds.ac.uk/464/)**    

## Licensing Information

Original dataset [license](https://archive.researchdata.leeds.ac.uk/464/): **Creative Commons Attribution 4.0 International (CC BY 4.0)**

### Contributions

 Original paper authors: Alqahtani, Mohammad and Atwell, Eric (2018) Annotated Corpus of Arabic Al-Quran Question and Answer. University of Leeds. https://doi.org/10.5518/356