arrow left facebook twitter linkedin medium menu play circle

Don’t Sleep on Sleeper Cells: How Big Data Drives Detection

October 5, 2017

DataVisor will be presenting tomorrow at the 2017 Grace Hopper Celebration in Orlando, Florida, the world’s largest gathering of women technologists produced by Our algorithm team technical lead, Hongyu Cui, Ph.D., will be presenting on Friday, October 6th during the Cybersecurity and Fraud Detection presentations session on the topic of “Don’t Sleep on Sleeper Cells: How Big Data Drives Detection.” The overall session will start at 10:30 am.

In her 20-minute presentation, Hongyu will go over the sleeper cells phenomenon, a new prevalent attack technique used by cyber attackers. These sleeper cells are dormant user account armies created by fraudsters which pose as legitimate users waiting for the opportune time to attack an online service. Since they are deliberately aged to bypass traditional rules-based detection systems, these sleeper cell accounts are extremely difficult to detect. This changes with the maturity of Big Data technology infrastructure and breakthroughs in unsupervised machine learning technology, which can review millions of users and billions of events per hour to detect these sleeper cells by identifying patterns of correlated attributes.

By effectively identifying fraudulent accounts before they attack, our unsupervised approach reduces the window of exposure while upholding a good user experience. DataVisor has analyzed over 639 billion events and protected over two billion accounts to date. If you are interested in learning more, feel free to reach out for more information at