README.md

---
license: mit
---

# 📂 Complaint_and_Sections Dataset

The **Complaint_and_Sections** dataset is a legal dataset containing **47,329 rows** of real-world complaints paired with their corresponding **Indian Penal Code (IPC)** sections. It is designed to support NLP-based legal tools such as IPC section classifiers, legal assistants, and crime-matching systems.

---

## 📄 Dataset Structure

- **Format**: JSON
- **Split**: `train` only
- **Total Samples**: 47,329
- **Languages**: English
- **License**: MIT

| Column | Type   | Description |
|--------|--------|-------------|
| `id`   | string | Unique identifier for each record |
| `text` | string | Complaint description (may contain multilingual text) |
| `labels` | list of strings | List of applicable IPC sections (multi-label) |

---

## 🔗 Part of LawSight Project

This dataset is a core component of the **LawSight Project** – an AI-powered legal assistant that:

- Suggests applicable IPC sections from user complaints
- Finds similar past cases
- Drafts editable FIRs using interactive Q&A

🚀 GitHub Repository: [IPCSight – LawSight](https://github.com/riyaindap7/LawSight)  
🌐 Hosted Model: [`ri2000/IPC_matcher`](https://huggingface.co/ri2000/IPC_matcher)

---
## 💡 Use Cases

- 🔍 **IPC Section Prediction**
- ⚖️ **Legal Complaint Classification**
- 🧠 **Multilingual NLP Research in Law**
- 📊 **Crime Pattern Analysis & Case Similarity**

---

## 🔒 Access

This dataset is restricted to ensure responsible usage due to the legal and sensitive nature of the content.

👉 **[Request Access via Google Form](https://forms.gle/rjLkpDcbj2piY7KX8)**

Once approved, you can clone or download the dataset using Hugging Face’s `datasets` library.

---

## 👩‍💻 Developed By

**Riya,Sachi & Tanmay** 
With support from the **LawsSight Team**

> Built with ❤️ for societal impact.