Upcoming additions to Financial-News-Multisource (12 new datasets) — feedback welcome

#2
by Brianferrell787 - opened

Hi all — quick update on the Financial-News-Multisource dataset.

This corpus is actively maintained. I’m integrating 12 additional datasets that expand recency (through 2025), add finance-specific sources (with tickers and timestamps), broaden general-news context for market events, include social discussion signals, and deepen tech-company coverage. As always, the corpus will be deduplicated, schema-consistent (date, text, extra_fields), and aligned to our research-only usage terms.

Why these additions help

  • Recency & continuity: Yahoo Finance (ongoing 2025) + NYT 2000–present help keep coverage fresh and continuous.
  • Finance focus: Jeet’s U.S. financial news and FinSen add stock-linked news and sentiment metadata.
  • General news context: All-the-News and Frontpage broaden macro/political coverage impacting markets.
  • Social discussion: Reddit finance posts capture retail investor discourse for robustness.
  • Tech domain: HackerNoon significantly increases tech-company news volume in 2022–2023.

Planned additions (links):

  1. https://huggingface.co/datasets/bwzheng2010/yahoo-finance-data
  2. https://huggingface.co/datasets/AndyReas/frontpage-news
  3. https://github.com/felixdrinkall/financial-news-dataset
  4. https://www.kaggle.com/datasets/eaglewhl/finsen-financial-sentiment-dataset
  5. https://huggingface.co/datasets/emilpartow/reddit_finance_posts_sp500
  6. https://www.kaggle.com/datasets/aryansingh0909/nyt-articles-21m-2000-present
  7. https://www.kaggle.com/datasets/jeet2016/us-financial-news-articles
  8. https://components.one/datasets/all-the-news-2-news-articles-dataset
  9. https://www.kaggle.com/datasets/julianschelb/newsdata
  10. https://www.kaggle.com/datasets/jonasbecker98/news-website-articles
  11. https://www.kaggle.com/datasets/parsonsandrew1/nytimes-article-lead-paragraphs-18512017
  12. https://huggingface.co/datasets/HackerNoon/tech-company-news-data-dump

Feedback welcome: If you have suggestions (e.g., preferred filters, additional sources, field/schema tweaks), please comment here or open an issue. The goal is to keep this dataset broadly useful for non-commercial, research-only work in financial NLP.

Thanks for using and supporting the project!

Brianferrell787 pinned discussion
Brianferrell787 changed discussion status to closed
Brianferrell787 changed discussion status to open
Brianferrell787 changed discussion status to closed

Sign up or log in to comment