---
language:
- ja
license: mit
library_name: transformers
pipeline_tag: text-classification
base_model: rinna/japanese-roberta-base
tags:
- japanese
- ragebait
- social-media
- binary-classification
model-index:
- name: Japanese X Ragebait - Rinna RoBERTa Base
  results:
  - task:
      type: text-classification
      name: Binary text classification
    dataset:
      name: Japanese X Ragebait test split
      type: custom
    metrics:
    - type: accuracy
      value: 0.829
      name: Accuracy
    - type: precision
      value: 0.8309859154929577
      name: Precision (YES)
    - type: recall
      value: 0.826
      name: Recall (YES)
    - type: f1
      value: 0.8289984609861489
      name: Macro-F1
---

# Japanese X Ragebait - Rinna RoBERTa Base

This is the Rinna RoBERTa component of the three-model ragebait detector from **“From Detection to Characterization: A Large-Scale Study of Ragebait on Japanese X”** (WI-IAT 2026). It is a binary Japanese text classifier fine-tuned from [`rinna/japanese-roberta-base`](https://huggingface.co/rinna/japanese-roberta-base).

## Labels

- `NO` (`0`): non-ragebait
- `YES` (`1`): ragebait

The labels are research annotations and should not be treated as facts about a post or its author.

## Training

- Labeled split: 16,558 posts, balanced between `YES` and `NO`
- Train/validation partition: 14,902 / 1,656
- Held-out test split: 2,000 posts, balanced between `YES` and `NO`
- Epochs: 3
- Learning rate: `3e-5`
- Batch size: 32
- Maximum sequence length: 256
- Warmup ratio: 0.1
- Weight decay: 0.01
- Seed: 42
- Best checkpoint criterion: validation Macro-F1

The public dataset repository contains only Post IDs and labels, not redistributed post text: [ZhiyangQi/japanese-x-ragebait](https://github.com/ZhiyangQi/japanese-x-ragebait).

## Test results

| Metric | Value |
| --- | ---: |
| Accuracy | 82.90% |
| Precision (`YES`) | 83.10% |
| Recall (`YES`) | 82.60% |
| Macro-F1 | 82.90% |

The confusion matrix is `[[832, 168], [174, 826]]`, ordered as `NO`, `YES`. Full training history and unrounded values are in `training_results.json`.

## Usage

```python
import torch
from transformers import AutoModelForSequenceClassification, AutoTokenizer

model_id = "ZhiyangQi97/japanese-x-ragebait-rinna-roberta-base"
tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=False)
model = AutoModelForSequenceClassification.from_pretrained(model_id)

text = "分類したい日本語テキスト"
inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=256)
with torch.inference_mode():
    prediction = model(**inputs).logits.argmax(dim=-1).item()

print(model.config.id2label[prediction])
```

Install `transformers`, `torch`, and `sentencepiece` before loading the tokenizer.

## Limitations and intended use

This model is intended for non-commercial academic research on Japanese ragebait. The training labels were created with an LLM-assisted pipeline and may contain errors. Ragebait is contextual and intent-based; performance may degrade on other periods, platforms, languages, or domains. Do not use this model to profile, rank, target, identify, harass, moderate, or make consequential decisions about individual users.

## License

The base model is distributed under the MIT License. See the [base model card](https://huggingface.co/rinna/japanese-roberta-base) for its terms and documentation.

## Citation

```bibtex
@inproceedings{qi2026ragebait,
  title = {From Detection to Characterization: A Large-Scale Study of Ragebait on Japanese X},
  author = {Qi, Zhiyang and Ito, Kazuhiro and Chen, Jinghui and Nakamura, Hibiki and Chen, Zhangxuan and Murata, Erina and Chujyo, Masaki and Toriumi, Fujio},
  booktitle = {WI-IAT 2026},
  year = {2026}
}
```