Transcripts of speeches given by various US politicians during the 2020 US Presidential Election were scraped from rev.com. The dataset includes manually added information such as location and speech type. The license is GPL 2.
Use Cases
- Analyze political messaging and rhetoric based on speech text.
- Compare speech styles and topics across different speakers.
- Study the relationship between speech content and event type (e.g., campaign speech, debate).
- Map political discourse to geographic locations based on the location column.
- Train text classification models to categorize speech types.
Strengths
- Transcripts are sourced from a specific platform (rev.com), providing a consistent base.
- Additional contextual metadata (location, type) has been manually added.
Limitations
- Row count and dataset size are unknown, which may limit suitability assessment.
- Column-level documentation is absent; field semantics must be inferred after download.
Provenance
- Source
- rev.com
- Collection Method
- Scraped transcripts with manually added metadata.
- Time Range
- 2020 US Presidential Election period
- Geography
- United States