[P] Central Bank Monetary Policy Dataset - 12 banks, 5000+ documents, sentiment labels

2026-02-07 · Source: Machine Learning · Field: Finance & Economics — Economic Analysis & Policy, Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, quick

Summary

A new dataset of central bank communications has been released, featuring over 5,000 documents from 12 major central banks, including the Federal Reserve, European Central Bank, and Bank of England. This dataset includes policy statements, meeting minutes, and speeches, all annotated with sentence-level sentiment labels (hawkish, dovish, neutral). It also integrates key economic indicators such as interest rates, foreign exchange rates, GDP, and inflation, primarily sourced from FRED. The dataset aims to provide a comprehensive resource for analyzing monetary policy and is accessible via a dashboard at monetary.live and on Hugging Face.

Key takeaway

For financial analysts and NLP engineers researching macroeconomic policy, this dataset offers a unique resource to train models on central bank sentiment. You can leverage the pre-labeled communications and integrated economic indicators to build predictive models for market reactions or to track policy shifts more effectively.

Key insights

A new dataset provides labeled central bank communications and economic indicators for monetary policy analysis.

Principles

• Central banks offer extensive public data.
• Sentiment analysis can quantify policy stances.

Method

The dataset compiles central bank documents and economic indicators, applying sentence-level hawkish/dovish/neutral sentiment labels to communications.

In practice

• Analyze monetary policy trends.
• Develop NLP models for financial text.

Topics

• Central Bank Communications
• Monetary Policy Analysis
• NLP Sentiment Analysis
• Economic Indicators
• Financial Datasets

Best for: NLP Engineer, AI Scientist, Research Scientist, Data Scientist, AI Data Scientist, AI Researcher

Related on AIssential

• Efficient Financial Language Understanding via Distillation with Synthetic Data — Distillation with synthetic data, guided by clustering real examples, enables efficient financial NLP in low-resource settings.
• I Analyzed 10,000 Financial News Headlines to See If the Media Predicts the Market — Financial news sentiment challenges negativity assumptions, varies by source, and requires extensive data for market correlation.
• Released a free 9.8M doc Indic multilingual corpus — Hindi, Bengali, Tamil, Telugu + 7 more (CC0, HuggingFace) [P] — A new 9.8M document, 11-language Indic multilingual corpus is now freely available under CC0.
• A Chinese Education Broadcast Emotion Corpus — A new Chinese education broadcast emotion corpus aids emotion recognition research.
• ConsumerBR: A Large-Scale Corpus of Consumer Complaints in Brazilian Portuguese — ConsumerBR offers a large, anonymized corpus of Brazilian Portuguese consumer complaints for NLP research.
• Python Sentiment Analysis: From Simple Tools to BERT — Effective sentiment analysis requires matching the right Python tool to text complexity and business needs.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning.