COVID-19 Fake News Headlines with Binary Fact-Check Labels

A collection of COVID-19-related headlines and claims shared across the internet, each labeled for veracity. The dataset was published by Sumit Banik in response to research demand for a combined fake news resource. It contains a binary outcome column where 0 indicates a fake headline and 1 indicates a true one.

Use Cases

• Train binary text classifiers for fake news detection based on labeled headlines.
• Analyze linguistic patterns of COVID-19 misinformation based on the headline text.
• Benchmark model performance on a domain-specific misinformation task using the provided binary labels.

Strengths

• Dataset addresses a specific, high-demand research question for combined COVID-19 fake news data.
• Clear binary labeling scheme (0 for fake, 1 for true) for straightforward classification tasks.

Limitations

• Row count, file formats, and column-level documentation are unknown, limiting suitability assessment.
• Last update date is unknown; freshness unverified.
• Data may reflect temporal and source bias inherent to internet-sourced claims from the pandemic period.

Provenance

Source

Sumit Banik via paperswithcode

Collection Method

Likely gathered from internet sources; specific collection method is not detailed.