naidukr/opentriage-dataset

---
language:
- en
multilinguality:
- monolingual
size_categories:
- 10K<n<100K
source_datasets:
- original
task_ids:
- text-classification
- text-generation
tags:
- disaster-response
- ticket-triage
- emergency-management
- knowledge-distillation
license: cc-by-4.0
---

# OpenTriage Dataset

A comprehensive dataset for disaster ticket triage and emergency response classification. This dataset contains labeled tickets from various disaster scenarios including earthquakes, hurricanes, floods, and other emergency situations.

## Dataset Description

The OpenTriage Dataset is designed to train and evaluate models for automated ticket classification in emergency response systems. It includes real-world inspired tickets with multiple classification labels and emergency-related information.

### Dataset Features

- **Disaster Ticket Classification**: Multi-label classification of emergency tickets
- **Student Training Data**: Educational examples for model distillation
- **Real-World Scenarios**: Data inspired by actual emergency response situations
- **Comprehensive Annotations**: Rich metadata and classification labels

## Dataset Structure

### Main Datasets

1. **large_triage_dataset.json** (18.7 MB)
- Primary dataset with comprehensive ticket triage records
- Largest dataset with extensive examples
- Full feature set and detailed annotations

2. **large_triage_training_data.json** (4.6 MB)
- Training-focused subset optimized for model training
- Balanced class distribution
- Cleaned and processed entries

3. **large_student_training_data.json** (4.2 MB)
- Student training data for knowledge distillation
- Simplified annotations for learning
- Ideal for distilled model training

### Sample & Reference Data

4. **sample_triage_tickets.json** (37 KB)
- Representative sample of the dataset
- Quick exploration and testing
- Format reference

5. **triage_training_data.json** (46 KB)
- Small training dataset
- Proof-of-concept examples
- Development and testing

6. **student_training_data.json** (44 KB)
- Compact student dataset
- Educational examples
- Model distillation reference

### Configuration Files

- **disaster_ticket_triage_config.json**: Configuration schema and parameters for ticket triage

## Data Format

Each record typically contains:
- `ticket_id`: Unique identifier
- `description`: Ticket description or incident report
- `category`: Primary classification category
- `tags`: Multiple classification tags
- `priority`: Urgency level
- `disaster_type`: Type of disaster (earthquake, hurricane, flood, etc.)
- Metadata and additional structured fields

## Dataset Statistics

- **Total Records**: ~100,000+ tickets across all files
- **Disaster Types**: Multiple categories (earthquakes, hurricanes, floods, wildfires, etc.)
- **Languages**: English
- **License**: CC-BY-4.0

## Usage

### Load with HuggingFace Datasets

```python
from datasets import load_dataset

# Load the full dataset
dataset = load_dataset("naidukr/opentriage-dataset", split="train")

# Load specific files
large_dataset = load_dataset("naidukr/opentriage-dataset", data_files="large_triage_dataset.json")
training_data = load_dataset("naidukr/opentriage-dataset", data_files="large_triage_training_data.json")
```

### Load with Pandas

```python
import json
import pandas as pd

with open('large_triage_dataset.json', 'r') as f:
data = json.load(f)

df = pd.DataFrame(data)
```

## Applications

- Emergency response systems
- Disaster ticket classification
- Emergency management training
- Knowledge distillation for light-weight models
- Multi-label classification research
- Real-time incident categorization

## Dataset Splits

The dataset includes multiple versions optimized for different use cases:
- **Full Dataset**: Complete training corpus (large_triage_dataset.json)
- **Training Subset**: Balanced training set (large_triage_training_data.json)
- **Distillation Data**: For creating smaller models (large_student_training_data.json)
- **Samples**: Quick reference and format validation

## Citation

If you use this dataset, please cite:

```bibtex
@dataset{opentriage_dataset_2026,
title={OpenTriage: Disaster Ticket Triage Dataset},
author={OpenTriage Team},
year={2026},
url={https://huggingface.co/datasets/naidukr/opentriage-dataset}
}
```

## License

This dataset is licensed under **CC-BY-4.0** (Creative Commons Attribution 4.0 International License).

You are free to:
- Share — copy and redistribute the material in any medium or format
- Adapt — remix, transform, and build upon the material for any purpose, even commercially

Under the following terms:
- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made

## Contact & Support

For questions, issues, or contributions related to this dataset:
- Visit: https://huggingface.co/datasets/naidukr/opentriage-dataset
- Related Model: https://huggingface.co/naidukr/Opentriage

## Changelog

### v1.0 (2026-04-18)
- Initial dataset release
- Multiple dataset sizes for different use cases
- Comprehensive documentation and usage examples
- Git LFS support for large files