---
task_categories:
- retrieval-augmented-generation
language:
- en
tags:
- legal
- india
- bns
- fir
pretty_name: Bharatiya Nyaya Sanhita (BNS) Definitions Dataset
size_categories:
- 1K<n<10K
---

# Dataset Card for BNS Definitions Dataset

BHARATIYA NYAYA SANHITA (BNS) DEFINITIONS DATASET

## Dataset Details

### Dataset Description

THIS IS A CLEANED AND CURATED DATASET OF DEFINITIONS UNDER THE **BHARATIYA NYAYA SANHITA, 2023 (BNS)**  
BUILT FOR USE IN LEGAL RAG (RETRIEVAL AUGMENTED GENERATION) SYSTEMS.

THE DATASET INCLUDES OFFICIAL SECTION TITLES AND THEIR DETAILED LEGAL DEFINITIONS  
TO HELP YOUR MODELS UNDERSTAND THE NUANCES OF INDIAN PENAL LAW BETTER ⚖️📜

- **Curated by:** [Navaneeth. K]

### Dataset Sources 
https://upload.indiacode.nic.in/schedulefile?aid=AC_CEN_5_23_00049_202346_1719552320687&rid=1191
https://www.indiacode.nic.in
https://www.indiacode.nic.in/  
https://huggingface.co/datasets  
https://www.chatgpt.com  
https://claude.ai/

## Uses

This dataset is designed to **enhance legal retrieval systems**, especially for tasks involving:

- FIR generation
- Legal document classification
- Intelligent legal assistants
- Chatbots that answer Indian criminal law queries

### Direct Use

I'm using this dataset to boost a legal RAG pipeline (AFIRGen) that **retrieves accurate section references and definitions**  
from BNS and explains them in simple terms to users or systems generating legal outputs.

## Dataset Structure

The dataset follows a simple JSONL format like:

```json
{"section": "106", "title": "Abetment of offence", "definition": "A person abets an offence when..."}