TweetQA: Question Answering Dataset for Social Media Posts

A large-scale dataset for question answering over social media data, created to address real-time knowledge needs. It was developed by ucsbai and last updated on Hugging Face in January 2024. The dataset focuses on informal text from platforms like Twitter, contrasting with previous QA datasets built on formal sources like news and Wikipedia.

Use Cases

• Training QA models based on informal social media text mentioned in the description
• Benchmarking NLP systems on real-time event comprehension based on the dataset's focus
• Developing applications that rely on knowledge extraction from social media posts

Strengths

• Described as the first large-scale QA dataset for social media data
• Focuses on real-time events and news reported on social platforms

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data may reflect temporal or platform bias inherent to social media sources

Provenance

Source

ucsbai

Freshness

Last updated 2024-01-24 08:53:20