SourceScore
SourceScore sub-tool · 2 of 4

Modern Citation Reference

Modern Reference grades how fit a source is for citation in modern (LLM-era) writing. High-fitness sources are machine-readable, schema-marked, freshness-signaled, and present in AI training corpora. Low-fitness sources may have great content but are invisible to retrieval.

Ranking — 50 sources

  1. #1
    DOI (CrossRef Resolver)
    doi.org
    A+·98
  2. #2
    U.S. Securities and Exchange Commission
    sec.gov
    A+·95
  3. #3
    arXiv
    arxiv.org
    A+·95
  4. #4
    U.S. National Institutes of Health
    nih.gov
    A·94
  5. #5
    U.S. Census Bureau
    census.gov
    A·94
  6. #6
    U.S. Bureau of Labor Statistics
    bls.gov
    A·93
  7. #7
    Wikipedia (English)
    en.wikipedia.org
    A·92
  8. #8
    PubMed
    pubmed.ncbi.nlm.nih.gov
    A·92
  9. #9
    Semantic Scholar
    semanticscholar.org
    A·92
  10. #10
    U.S. Food and Drug Administration
    fda.gov
    A·92
  11. #11
    U.S. National Oceanic and Atmospheric Administration
    noaa.gov
    A·92
  12. #12
    Our World in Data
    ourworldindata.org
    A·90
  13. #13
    World Bank
    worldbank.org
    A·90
  14. #14
    Reuters
    reuters.com
    A·88
  15. #15
    World Health Organization
    who.int
    A·88
  16. #16
    Associated Press
    apnews.com
    A·86
  17. #17
    The Guardian
    theguardian.com
    A·86
  18. #18
    Pew Research Center
    pewresearch.org
    A·86
  19. #19
    International Monetary Fund
    imf.org
    A·86
  20. #20
    Nature
    nature.com
    A·85
  21. #21
    The Economist
    economist.com
    A·85
  22. #22
    ProPublica
    propublica.org
    B·84
  23. #23
    Science
    science.org
    B·84
  24. #24
    New England Journal of Medicine
    nejm.org
    B·84
  25. #25
    BBC News
    bbc.com
    B·83
  26. #26
    The New York Times
    nytimes.com
    B·82
  27. #27
    Encyclopædia Britannica
    britannica.com
    B·82
  28. #28
    The Lancet
    thelancet.com
    B·82
  29. #29
    The Washington Post
    washingtonpost.com
    B·81
  30. #30
    NPR
    npr.org
    B·80
  31. #31
    Ars Technica
    arstechnica.com
    B·80
  32. #32
    STAT News
    statnews.com
    B·80
  33. #33
    MIT Technology Review
    technologyreview.com
    B·80
  34. #34
    Wired
    wired.com
    B·80
  35. #35
    Al Jazeera English
    aljazeera.com
    B·80
  36. #36
    Financial Times
    ft.com
    B·78
  37. #37
    The Wall Street Journal
    wsj.com
    B·78
  38. #38
    Foreign Affairs
    foreignaffairs.com
    B·78
  39. #39
    Politico
    politico.com
    B·78
  40. #40
    404 Media
    404media.co
    B·76
  41. #41
    Bloomberg News
    bloomberg.com
    B·75
  42. #42
    TechCrunch
    techcrunch.com
    B·70
  43. #43
    The Verge
    theverge.com
    B·70
  44. #44
    Medium
    medium.com
    C·65
  45. #45
    Forbes
    forbes.com
    C·65
  46. #46
    Fox News
    foxnews.com
    C·65
  47. #47
    HuffPost
    huffpost.com
    C·65
  48. #48
    Statista
    statista.com
    C·56
  49. #49
    BuzzFeed
    buzzfeed.com
    F·38
  50. #50
    Daily Mail
    dailymail.co.uk
    F·30

How Modern Reference is scored

Modern Reference combines four signals: presence in major LLM training corpora, structured-data quality (Article + Organization + DefinedTerm schema, JSON-LD), freshness signals (datePublished + dateModified, last-verified visibility), and machine-readability (DOIs, stable URLs, full-text search APIs).

Government primary sources with public APIs score highest. User-generated platforms with mixed-quality posts score lowest because retrieval models down-weight them after the 2024 Helpful Content shifts.