---
title: GDPR Cases Demo
emoji: 🏛️
colorFrom: blue
colorTo: green
sdk: docker
app_file: app.py
pinned: false
---

# GDPR Cases - Interactive Demo

An interactive Streamlit application for exploring GDPR formalization cases and understanding step-by-step rule evaluation using the Pythen framework.

## Features

- **Dataset Browsing**: Browse all 60 verified GDPR cases from the dataset
- **Sample Selection**: Choose any sample by article and ID
- **Scenario Viewing**: Read the complete legal scenario for each case
- **Facts Display**: View extracted atomic facts used in evaluation
- **Rule Tree Visualization**: Inspect the formal rule tree in JSON format
- **Quality Metrics**: See evaluation scores from multiple verifiers
- **Step-by-Step Evaluation**: Understand how Pythen evaluates rules
- **Ground Truth**: View the expected legal outcome for each case

## Dataset

This demo uses the **GDPR Cases** dataset (`nguyenthanhasia/gdpr-cases`):
- **60 verified samples** of GDPR formalization cases
- **11 columns** including scenario, rule tree, facts, and evaluation scores
- **High quality**: All samples verified by legal experts

## How It Works

### Rule Evaluation Process

1. **Parse Rule Tree**: Hierarchical structure of conditions and predicates
2. **Extract Facts**: Atomic facts from the legal scenario
3. **Traverse Tree**: Evaluate from root through all nodes
4. **Apply Operators**:
   - `ANY`: At least one condition must be true
   - `ALL`: All conditions must be true
5. **Derive Label**: Final boolean result (TRUE/FALSE)

### Pythen Framework

Pythen is a formal representation framework for legal rules that:
- Separates conditions from exceptions
- Uses logical operators (ANY, ALL) for composition
- Enables automated evaluation of legal provisions
- Supports complex nested rule structures

## Citation

If you use this demo or dataset, please cite:

```bibtex
@article{nguyen2026gdpr,
  title={GDPR Auto-Formalization with AI Agents and Human Verification},
  author={Nguyen, Ha Thanh and Fungwacharakorn, Wachara and Wehnert, Sabine and Zin, May Myo and Kong, Yuntao and Xue, Jieying and Araszkiewicz, Michał and Goebel, Randy and Satoh, Ken},
  journal={arXiv preprint arXiv:2604.14607},
  year={2026}
}
```

## Related Resources

- [Dataset on Hugging Face](https://huggingface.co/datasets/nguyenthanhasia/gdpr-cases)
- [Paper on arXiv](https://arxiv.org/abs/2604.14607)