Twitter Financial News Sentiment: 11,932 Annotated Market Tweets

11,932 English-language finance tweets annotated for sentiment classification, released by zeroshot and last updated in February 2024. The corpus categorizes social media posts into three distinct market outlooks: Bearish, Bullish, and Neutral.

Use Cases

• Training sentiment analysis models to distinguish between Bullish and Bearish market signals in social media text
• Evaluating NLP performance on finance-specific terminology and ticker symbols
• Analyzing the distribution of Neutral vs. directional sentiment in financial news tweets

Strengths

• 11,932 annotated documents
• Three-class sentiment labels specifically for financial markets
• Domain-specific focus on financial news rather than general social media

Limitations

• Limited to three sentiment categories which may miss nuanced market signals
• Potential for temporal staleness as financial slang and hashtags evolve
• Geographic bias towards US markets as indicated by metadata tags

Provenance

Source

Twitter API

Collection Method

Scraped via API and annotated

Freshness

Last updated February 2024; reflects historical Twitter data.

Geography

United States