Home / Market & alternative data
bardsai/finance-sentiment-ja-base
Three-way sentiment labels for Japanese financial text
What it is
A sequence classifier built on a Japanese BERT base model for the sentiment of financial news. Given a sentence, it returns one of three labels, positive, negative, or neutral, and the intended call path is the transformers sentiment-analysis pipeline. Researchers and developers who want a tone tag on Japanese company news are the audience. Training used a translated Financial PhraseBank rather than original Japanese market text, and the inherited ShareAlike license requires further derivatives to stay under a compatible ShareAlike term.
At a glance
Research onlyOur rating, based on popularity, maintenance and how ready it is for real use.
| Best for | Professional quants |
|---|---|
| Used for | Data analysis, Strategy research |
| Markets | Multi-market |
| Stack | Python |
| Learning curve | Easy to start |
| Practical value | Medium practical value |
| Cost | Free and open source |
| Hardware | GPU optional |
GitHub stars, last 30 days
Daily snapshots since 2026-09-28 (up to 30 days): +104 over the period, now 876. Gaps mean no snapshot was taken that day.
text-classification · transformers, pytorch, safetensors, bert, text-classification, financial-sentiment-analysis
Similar tools
Portfolio analytics and performance tear sheets for quants
Financial function library for quantitative analysis in Python
Three-way sentiment labels for Chinese financial text
Jupyter notebooks companion to Machine Learning for Trading
Additive time series forecasting with automatic seasonality handling
scikit-learn compatible Python library for portfolio optimization