Sign in to view source links and access this dataset
Description
US election-related tweets from the 2024 cycle, cleaned and converted to Parquet format. The dataset is organized into parts, with each part containing 1,000,000 tweets and each chunk file containing 50,000 tweets. It was uploaded by the author 'deadbirds' to Hugging Face and last updated on May 31, 2025.
Use Cases
Analyze public sentiment and political discourse based on election-related tweet content.
Study information diffusion and network dynamics based on the social media data.
Train or fine-tune language models for political text classification based on the tweet corpus.
Strengths
Each part contains a large volume of 1,000,000 tweets.
Data has been cleaned and converted to the efficient Parquet file format.
The dataset is structured into manageable chunks of 50,000 tweets each.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
The total number of rows/parts is unknown, which may limit suitability assessment.
Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
Source
USC (University of Southern California) X 24 US Election Twitter/X Dataset.
Collection Method
Likely collected from the Twitter/X platform API, then cleaned and converted.
Time Range
Related to the 2024 US elections.
Freshness
Last updated 2025-05-31 19:13:04; freshness should be verified.
Geography
United States, inferred from the topic.
License is unknown; users must verify permissions before use.