Skip to content

Dataset · takala

Financial sentences with sentiment

takala/financial_phrasebank

Around 5k English financial sentences labeled as positive, negative, or neutral.

What's inside

A collection of approximately 5,000 English sentences manually annotated for sentiment: positive, negative, or neutral. Annotators evaluated sentences from an investor's perspective and the news's impact on stock prices. The data relates to finance and economics.

What you can do with it

  • Train or fine-tune a financial sentiment classifier
  • Test LLMs on financial news classification
  • Create a benchmark for financial sentiment analysis
Tasks
Text classification
Languages
English
Size
1K–10K rows
Downloads this month
6.2K
Likes
292

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("takala/financial_phrasebank")