Rieth Research

Central Bank Communication Analytics & Financial NLP Research

Rieth Research specializes in natural language processing of central bank communications and monetary policy analysis. We develop domain-specific AI models and datasets to advance understanding of financial text patterns and policy communication strategies.

Research Focus

• Central Bank Communication Analysis: NLP techniques for analyzing statements from RBNZ, Federal Reserve, and other central banks
• Monetary Policy Stance Detection: Automated classification of hawkish/dovish/neutral policy language
• Financial Domain NLP: Custom tokenizers and models optimized for financial and economic text
• Time-Series Text Analytics: Tracking communication patterns and language evolution over time

Available Models

Classification Models

• Central Bank Classifier: Identifies which central bank issued a communication
• Policy Stance Analyzer: Detects the monetary policy orientation (hawkish/dovish/neutral)
• Communication Type Classifier: Distinguishes between statements, implementation notes, and other documents

Analysis Tools

• Financial Domain Tokenizer: Custom tokenizer trained on central bank communications
• Temporal Analysis Pipeline: Tracks language changes across time periods
• Policy Language Extractor: Identifies key monetary policy terms and phrases

Datasets

Central Bank Communications

• RBNZ OCR Statements: Official Cash Rate decisions (2006-2012)
• Federal Reserve FOMC: Statements and implementation notes (2012-2017)
• Annotated Policy Data: Labeled datasets for training and evaluation

Methodology

Our approach combines:

• Transformer-based models (BERT, RoBERTa) fine-tuned on financial text
• Domain-specific preprocessing for OCR-corrected documents
• Time-series analysis for tracking communication evolution
• Rigorous evaluation with financial domain-specific metrics

Mission

"Democratizing access to advanced NLP tools for financial research through open-source models and datasets."

Contact & Collaboration

For research collaborations, dataset contributions, or model inquiries:

• GitHub: Central Bank Text Analytics Repository
• Email: research@riethresearch.com
• Hugging Face: @rieth-research

Contributing

We welcome contributions to our models and datasets. Please see our contribution guidelines for:

• Data preprocessing standards
• Model training protocols
• Evaluation metrics
• Documentation requirements

Citation

If you use our models or datasets in your research, please cite:

@misc{rieth-research-2026,
  author = {Rieth Research},
  title = {Central Bank Communication Analytics Models},
  year = {2026},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/rieth-research}}
}

---

Advancing understanding of central bank communications through advanced NLP techniques