---
pretty_name: "UAE Sales Table QA (Arabic)"
license: cc-by-4.0
language:
- ar
task_categories:
- question-answering
- table-question-answering
tags:
- pandas
- arabic
- table-reasoning
- data-analysis
- uae
size_categories:
- 1K<n<10K
---

# 🇦🇪 UAE Sales Table QA (Arabic)

> ⚠️ **Note:** All data in this dataset is **synthetically generated** using random values for learning and experimentation purposes. It does **not** represent real-world business data.

## 🧠 Overview
**UAE Sales Table QA (Arabic)** is an Arabic **Question–Answering dataset for table reasoning and data analysis**, generated from **21** UAE-style **CSV** tables.  
Each example includes:

- **Question** — a natural-language query about the data  
- **Steps** — human-readable reasoning steps describing the logic used to compute the answer  
- **Answer** — the numeric result automatically calculated using pandas  
- **Source_File** — the CSV file the example was computed from, for traceability  

This dataset covers operations like filtering, grouping, and aggregation, all verified using **pandas**.

---

## 📂 Files
| File | Description |
|------|-------------|
| `uae_qa_sales.json` | Full Q&A dataset (includes `Source_File` for linking) |
| `csvs/` | Folder containing the 21 source **CSV** files referenced by `Source_File` |

**Total Q&A items:** ~6,300

---

## 🧩 Schema
```json
{
  "Question": "ما هو متوسط الإيراد في دبي؟",
  "Steps": [
    "1. فلتر الصفوف بحيث تكون المدينة = 'دبي'.",
    "2. اختر عمود 'الإيراد'.",
    "3. احسب المتوسط."
  ],
  "Answer": "2766.62",
  "Source_File": "csvs/sales_uae_7.csv"
}
```

---

## 💻 How to Load
```python
from datasets import load_dataset

ds = load_dataset("Synctic/uae-sales-table-qa", data_files="uae_qa_sales.json")
print(len(ds["train"]))          # number of examples
print(ds["train"][0])            # first record
```

---

## ⚙️ Creation Process
1. Multiple **synthetic CSV** tables were generated with random UAE-style sales data (cities, products, reps, dates).  
2. Ten base question templates covered averages, sums, min/max, and grouping logic; an additional template combined (city × product × year) to expand variations.  
3. Each CSV produced multiple Q&A variations.  
4. All answers were computed using **pandas** for accuracy.  
5. The **`Source_File`** field links every Q&A to its corresponding CSV file in `csvs/`.

---

## 📊 Dataset Stats
| Property | Value |
|-----------|--------|
| **Language** | Arabic 🇦🇪 |
| **Domain** | Sales / Business Analytics |
| **Entries** | ~6,300 |
| **Source CSVs** | 21 |
| **Format** | JSON (UTF-8, human-readable) |

---

## 🔖 License
**CC-BY-4.0** — You are free to use, modify, and share this dataset with proper attribution.

---

## 📬 Citation
If you use or reference this dataset, please cite it as:

```
@dataset{AlYazidi2025_UAE_Sales_Table_QA,
  author = {Saif Al-Yazidi},
  title  = {UAE Sales Table QA (Arabic)},
  year   = {2025},
  url    = {https://huggingface.co/datasets/Synctic/uae-sales-table-qa}
}
```

---

## 🌟 Author & Acknowledgment
Created by **Saif Al-Yazidi (Synctic)** as a personal learning project on:
- **pandas** data manipulation  
- **CSV** data generation  
- **Arabic AI reasoning dataset creation**  

This project connects structured tabular data with natural-language reasoning — a foundation for future **Arabic AI reasoning** and **data analysis** research.