4Chan Pol: Sanitized Text Data from an Online Forum

u84u's dataset is a sanitized version of text data from the 'Politically Incorrect' (/pol/) board of the 4chan forum. The data was originally extracted from the 'ermcy/pol' dataset hosted on HuggingFace, which covers posts from June to November 2016. The dataset was processed and re-hosted by u84u, with a last recorded update in May 2026.

Use Cases

• Analyze linguistic patterns and rhetoric based on forum text data.
• Train or benchmark models for toxic language detection based on content from an anonymous platform.
• Study the evolution of topics and narratives over time within the described date range.
• Conduct social network analysis based on user interaction patterns suggested by the forum structure.

Strengths

• The dataset is derived from a known academic source, with the original paper cited (arXiv:2001.07487).
• The processing pipeline is described, including steps for extraction and decompression.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Original data from the 4chan /pol/ board, via the 'ermcy/pol' dataset on HuggingFace.

Collection Method

Initial extraction from a hosted tarball, followed by decompression and conversion to parquet format.

Time Range

Likely covers June to November 2016, as indicated by the source filename 'pol_0616-1119_labeled'.

Freshness

Last updated 2026-05-10 16:02:09; freshness should be verified.