Cyberanalytics: Protecting us with high-performance computing
When we think of cybersecurity, we often think of the outsider trying to hack into our computer systems. But, another challenge is how we identify and defend against an insider, oftentimes a lone wolf, who knows our procedures and safety precautions.
If we want to protect ourselves from both scenarios, we must increase our reliance on high-performance computing, especially the graph analytic research we conduct at Georgia Tech, says David Bader, chair of the School of Computational Science and Engineering in the College of Computing.
Graphs help us discover patterns and relationships hidden in massive amounts of data. These graphs are comprised of interconnected vertices (nodes) and lines (edges), and these graphs change over time.
In the realm of cybersecurity, the vertices are people, places, and things, and the edges represent their interactions. By designing fast, using theoretic algorithms on large-scale graphs, we can produce insights in near-real time. This is crucial because cybersecurity analysts often are overwhelmed with thousands of alerts to review, and our algorithms may direct them immediately to the most important ones.
We leave a digital trace every time we use a key card to get through a door, log in to a computer, or send an email. Security officers need to analyze this information so they can understand our patterns and identify potential threats.
These massive-scale datasets are often unstructured and challenging to inspect. The emerging graph technology we are developing at Georgia Tech has the potential to be the best and most efficient way to prevent future attacks where we work and live, says Bader.
—David Bader chairs the School of Computational Science and Engineering in the College of Computing.