README.md

---
language:
- ar
license: apache-2.0
task_categories:
- multiple-choice
- question-answering
pretty_name: Arabic Accounting MCQ Training Dataset
tags:
- accounting
- mcq
- arabic
- training
- education
dataset_info:
  features:
  - name: id
    dtype: string
  - name: query
    dtype: string
  - name: answer
    dtype: string
  - name: text
    dtype: string
  - name: choices
    list: string
  - name: gold
    dtype: int64
  - name: conversations
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  splits:
  - name: train
    num_bytes: 853344
    num_examples: 249
  download_size: 251153
  dataset_size: 853344
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---

# Arabic Accounting MCQ Training Dataset

Training dataset for Arabic accounting multiple choice questions with English letter choices.

## Dataset Structure

- **Format**: Multiple choice questions (4 options)
- **Language**: Arabic questions with English letter choices
- **Domain**: Accounting and finance
- **Size**: ~80% of total dataset

## Fields

- `id`: Unique identifier
- `query`: Full MCQ prompt with instructions
- `answer`: Correct answer letter (a, b, c, d)
- `text`: Question text without instructions
- `choices`: List of options ['a', 'b', 'c', 'd']
- `gold`: Zero-based index of correct answer (0-3)

## Example

```json
{
  "id": "accounting_mcq_00001",
  "query": "اقرأ السؤال التالي بعناية واختر الإجابة الصحيحة...",
  "answer": "d",
  "text": "السؤال: [accounting question]...",
  "choices": ["a", "b", "c", "d"],
  "gold": 3
}
```

## Usage

```python
from datasets import load_dataset

dataset = load_dataset("SahmBenchmark/arabic-accounting-mcq_train")
train_data = dataset['train']

for example in train_data:
    print(f"Question: {example['text']}")
    print(f"Choices: {example['choices']}")
    print(f"Answer: {example['answer']}")
```

For evaluation data, see: `SahmBenchmark/arabic-accounting-mcq_eval`