Data sources
How MysteryAtlas turns sources into mystery case files.
MysteryAtlas is built around source inspection. Cases are generated from public records, original reporting, reference archives, video and audio metadata, structured databases, and community signals, then organized into evidence, debunks, timelines, and current assessments.
Official records and public archives
Government and institutional records carry high evidentiary weight when available. Examples include AARO UAP releases, NASA material, National Archives records, FBI Vault files, CIA Reading Room documents, military reports, court records, and aviation safety records.
Reference archives and Wikipedia trails
Wikipedia is treated as a discovery and orientation layer, not as final evidence. MysteryAtlas uses article references, external links, dates, aliases, and historical context to find source pages that can be inspected directly.
Publisher reporting
News articles, local reporting, documentaries, interviews, museum pages, and specialist publications help reconstruct what was reported at the time and how interpretations changed later.
Structured mystery databases
Structured sources such as NUFORC, MUFON, BFRO, public sighting records, case catalogs, and research databases help discover clusters, recurring descriptions, locations, dates, and witness patterns.
Video, audio, and media metadata
YouTube pages, documentary listings, podcast archives, video embeds, thumbnails, transcripts, and audio metadata can provide useful media provenance when they point back to a specific report or case claim.
Community and discussion signals
Reddit threads, forums, social posts, and user submissions can surface aliases, trend direction, and public interest. They are low-trust signals until matched with independent reporting or primary records.
Source processing workflow
- 01Discover source URLs through fixed feeds, sitemaps, RSS, search discovery, Wikipedia references, public databases, and case-specific source lists.
- 02Store each page, document, report, video entry, or forum thread as a raw item with publisher, date, title, URL, language, region, and access status.
- 03Classify whether the item is relevant to UFO or UAP cases, cryptids, unexplained events, strange signals, missing cases, ancient mysteries, government projects, or Chinese mystery reports.
- 04Cluster related raw items into case candidates using event name, aliases, location, date, source references, witness details, and similarity signals.
- 05Extract evidence, explanations, limitations, short original excerpts, source summaries, and Atlas interpretation while keeping the original publisher visible.
- 06Publish or update a case only when the source set is strong enough for the page type; lower-confidence material can remain internal or appear as developing coverage.
What counts as evidence?
A source matters more when it has a visible publisher, a date, original reporting, documents, images, video, audio, witness records, official statements, or direct observations. Search engines, social posts, and aggregation pages help discovery, but they do not replace the publisher, archive, or primary document.
Copyright and excerpts
MysteryAtlas does not republish full articles, paywalled works, or copyrighted media. Case pages may show short attributed excerpts, source summaries, and links so readers can inspect why a source is relevant. Copyright remains with the original authors, publishers, archives, and rights holders.
AI-assisted analysis
AI helps classify sources, cluster related reports, extract evidence, identify debunks, summarize original material, and draft current assessments. These outputs are editorial aids, not primary evidence. Readers should open source links and compare independent records before relying on any claim.
Corrections, missing source suggestions, media provenance issues, and copyright concerns can be sent to service@mysteryatlas.net.