Datasets:

golfplan18
/

msi-corpus

license: cc0-1.0
language:
  - en
pretty_name: Main Street Independent Corpus
task_categories:
  - text-generation
  - text-classification
  - summarization
tags:
  - news
  - journalism
  - ai-generated
  - cc0
size_categories:
  - 10K<n<100K
configs:
  - config_name: default
    data_files:
      - split: news
        path: articles.jsonl
      - split: opinion
        path: columns.jsonl

Main Street Independent Corpus

The complete Main Street Independent archive as open data: 17,825 news articles and 3,295 opinion columns, each with full text and metadata. AI-generated news and opinion, dedicated to the public domain under CC0 — no rights reserved. Refreshed daily.

Load it

from datasets import load_dataset
ds = load_dataset("golfplan18/msi-corpus")            # 'news' and 'opinion' splits
print(ds["news"][0]["headline"], ds["news"][0]["text"][:200])

Fields

slug, url, headline, publish_date, section, primary_entities, primary_themes, topic_tags, license, text (and pen_name for opinion columns).

Also available

Bulk Markdown + richer JSONL indexes (atomic claims, sources): https://github.com/Golfplan18/msi-corpus
Natural-language search + MCP endpoint: https://mainstreetindependent.com/data

License

CC0 1.0 Universal. Public domain — use freely, no attribution required.