This repository provides public research materials related to sentiment analysis of FOMC minutes and central bank communication.
The related FinBERT model for FOMC aspect sentiment classification is available on Hugging Face:
https://huggingface.co/Wonseong/FinBERT-FOMC-aspects
https://huggingface.co/ZiweiChen/FinBERT-FOMC
The model is built for text classification across FOMC-relevant aspects such as growth, employment, and inflation.
This work is connected to research on financial sentiment analysis and central bank communications, including:
-
FinBERT-FOMC: Fine-tuned FinBERT Model with Sentiment Focus Method for Enhancing Sentiment Analysis of FOMC Minutes https://dl.acm.org/doi/abs/10.1145/3604237.3626843
-
Is Small Really Beautiful for Central Bank Communication? Evaluating Language Models for Finance: Llama-3-70B, GPT-4, FinBERT-FOMC, FinBERT, and VADER https://dl.acm.org/doi/10.1145/3677052.3698675
-
Analyzing FOMC Minutes: Accuracy and Constraints of Language Models https://arxiv.org/abs/2304.10164
Large or working data files are not stored in this repository. Use the linked Hugging Face model page for model access and related resources.
If you use these materials, please cite the related papers above and reference the linked Hugging Face model where appropriate.