Anomaly Detection at Multiple Scales
{{Update|date=October 2021}}
{{Infobox
|name =
|title = Anomaly Detection at Multiple Scales
|label2 = Establishment
|data2 = 2011
|label3 = Sponsor
|data3 = DARPA
|label4 = Value
|data4 = $35 million
|label5 = Goal
|data5 = Detect insider threats in defense and government networks
|label7 = Website
|data7 = [http://www.darpa.mil/Our_Work/I2O/Programs/Anomaly_Detection_at_Multiple_Scales_(ADAMS).aspx www.darpa.mil]
}}
Anomaly Detection at Multiple Scales, or ADAMS was a $35 million DARPA project designed to identify patterns and anomalies in very large data sets. It is under DARPA's Information Innovation office and began in 2011{{cite web|url=http://www.darpa.mil/Our_Work/I2O/Programs/Anomaly_Detection_at_Multiple_Scales_(ADAMS).aspx|title=ADAMS|publisher=DARPA Information Innovation Office|accessdate=2011-12-05|archive-url=https://web.archive.org/web/20120121113838/http://www.darpa.mil/Our_Work/I2O/Programs/Anomaly_Detection_at_Multiple_Scales_(ADAMS).aspx|archive-date=2012-01-21|url-status=dead}}{{cite news|url=https://www.fbo.gov/download/2f6/2f6289e99a0c04942bbd89ccf242fb4c/DARPA-BAA-11-04_ADAMS.pdf|title=Anomaly Detection at Multiple Scales (ADAMS) Broad Agency Announcement DARPA-BAA-11-04|publisher=General Services Administration|date=2010-10-22|accessdate=2011-12-05|archive-url=https://web.archive.org/web/20120406072850/https://www.fbo.gov/download/2f6/2f6289e99a0c04942bbd89ccf242fb4c/DARPA-BAA-11-04_ADAMS.pdf|archive-date=2012-04-06|url-status=dead}}{{cite magazine|url=https://www.wired.com/dangerroom/2010/10/darpa-starts-sleuthing-out-disloyal-troops/|title=Darpa Starts Sleuthing Out Disloyal Troops|first=Spencer|last=Ackerman|magazine=Wired|date=2010-10-11|accessdate=2011-12-06}}{{cite news|url=https://edition.cnn.com/2010/US/10/27/pentagon.e.mail.profiling/|title=Military wants to scan communications to find internal threats|first=Charley|last=Keyes|work=CNN|date=2010-10-27|access-date=2011-12-06}} and ended in August 2014{{cite web | url=https://apps.dtic.mil/sti/citations/ADA625184 | title=DARPA ADAMS Project }}
The project was intended to detect and prevent insider threats such as "a soldier in good mental health becoming homicidal or suicidal", an "innocent insider becoming malicious", or "a government employee
A final report was published on May 11, 2015, detailing a system known as Anomaly Detection Engine for Networks, or ADEN, developed by the University of Maryland, College Park, whose goal was to "identify malicious users within a network." Using multiple datasets from Wikipedia, Slashdot, and others, researchers were able to identify vandals and malicious users on a website using both conventional algorithms and artificial intelligence.{{Cite web |last=Subrahmanian |first=V. S. |date=11 May 2015 |title=Final Report for the DARPA ADAMS Project |url=https://apps.dtic.mil/sti/pdfs/ADA625184.pdf |access-date=4 August 2024}}
The Proactive Discovery of Insider Threats Using Graph Analysis and Learning was part of the ADAMS project.{{cite news|url=http://www.foxnews.com/scitech/2011/12/03/could-us-government-start-reading-your-emails/|archive-url=https://web.archive.org/web/20111203222717/http://www.foxnews.com/scitech/2011/12/03/could-us-government-start-reading-your-emails/|url-status=dead|archive-date=December 3, 2011|title=Could the U.S. Government Start Reading Your Emails?|first=John|last=Brandon|publisher=Fox News|date=2011-12-03|accessdate=2011-12-06}} The Georgia Tech team includes noted high-performance computing researcher David Bader (computer scientist).{{cite web|url=http://www.cse.gatech.edu/taxonomy/term/153/all|title=Anomaly Detection at Multiple Scales|publisher=Georgia Tech College of Computing|accessdate=2011-12-06}}