Spotlight on New Propositions - Cyber

Cyber Security threats to
Open Government Data
Vishal Marria
April 2014
The Govt Cyber threat Landscape : Be Afraid
The US Department of Defense detected 300 Million network
penetration attempts last year, up from 6 million in Yr. 2000
The US Govt Accountability Office found in its survey that 23 of
the 24 agencies surveyed did not have computer security
protections in place
Govt computer networks are constantly bombarded with cyber
threats from sources around the world
It is evidenced that, Govt departments sharing open data
attracts more global attention on cyber threats
The analytics approach to Cyber security….
Case Study: Undisclosed Govt Organization
• A government organization needed to coordinate and monitor the
defense of its vast global networks
• 90 percent of the alerts were of a probing nature that did not require
immediate action, a lot of valuable time was wasted on unnecessary
• In addition, many network monitors produced huge volumes of data in
different formats, and the dept did not have any way of doing historical
analysis of data visualization on this massive threat data
Case Study: The challenge
The organization asked SAS, “Can you…”
Give us a real-time network defense system that automatically generates
attack alerts when threat response actions are required, while also
dramatically reducing the number of false positives?
Pinpoint anomalous properties of network traffic, which would normally go
undetected, as well as uncover hidden relationships and behavior patterns
Aggregate, correlate and merge data from all our network monitoring devices
and any other data sources so we could achieve complete network domain
Case Study: The solution
SAS developed a hybrid cyber security defense mechanism due to
the global nature of access to the govt organization’s web channels
Case Study: The results
Potential threats were recognized sooner than ever before
Analysts were able to investigate incidents on a network view of
data at greater speed
Billions of rows, rapidly explored in seconds . Scalability to handle
growing volumes of data
Critical pain point of “Alert Prioritization” achieved
Thank you

similar documents