thoriqtau/ner-kata-kasar

NER Kata Kasar - Indonesian Offensive Language Named Entity Recognition Dataset

Dataset Description

This dataset is designed for Named Entity Recognition (NER) tasks specifically focused on identifying and classifying offensive/inappropriate words in Indonesian text. The dataset contains labeled tokens that can be used to train models for detecting various types of offensive language patterns.

Dataset Details

• Language: Indonesian (Bahasa Indonesia)
• Task: Token Classification / Named Entity Recognition
• License: MIT
• Format: CSV

Label Schema

The dataset uses the following labeling scheme for offensive language classification:

Label	Description
<PAD>	Padding token
O	Other tokens (non-entity)
FIS-NEG	Physical offensive words (negative)
SEKS-NEG	Sexual offensive words (negative)
PER-NEG	Personal attack offensive words (negative)
FIS-POS	Physical words with positive meaning
HWN-POS	Animal references with positive meaning
HWN-NEG	Animal references with negative meaning

Usage

Loading the Dataset

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("thoriqtau/ner-kata-kasar")

# View sample data
print(dataset["train"][0])

# Access the training split
train_data = dataset["train"]
print(f"Number of examples: {len(train_data)}")

Data Format

Each example in the dataset contains:

• tokens: List of tokenized words
• labels: Corresponding NER labels for each token

Example Usage for Model Training

from datasets import load_dataset

dataset = load_dataset("thoriqtau/ner-kata-kasar")

print(dataset["train"][0])

Applications

This dataset can be used for:

• Content Moderation: Automatically detect offensive language in Indonesian text
• Social Media Monitoring: Filter inappropriate content on platforms
• Text Classification: Build robust offensive language detection systems
• Research: Study patterns of offensive language in Indonesian

Ethical Considerations

This dataset contains offensive language examples for research and content moderation purposes. Users should:

• Use this dataset responsibly and ethically
• Consider the potential impact of automated offensive language detection
• Be aware that language and cultural contexts may vary
• Implement appropriate safeguards when deploying models trained on this data

Limitations

• The dataset is focused on Indonesian language only
• Labels are based on specific cultural and linguistic contexts
• May not cover all possible forms of offensive language
• Regular updates may be needed to capture evolving language patterns

Contributing

If you find issues with the dataset or have suggestions for improvements, please feel free to reach out or contribute to the dataset maintenance.

Citation

If you use this dataset in your research, please cite:

@dataset{thoriqtau_ner_kata_kasar_2025,
  author = {Zaini Thoriq Taufiqurrahman},
  title = {NER Kata Kasar: Indonesian Offensive Language Named Entity Recognition Dataset},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/thoriqtau/ner-kata-kasar}
}

Contact

For questions or feedback about this dataset, please contact the author or create an issue in the dataset repository.