Network Traffic Analysis with Query Driven Visualization Session: HPC Analytics Challenge Finalists: Session 1 Event Type: HPC Analytics Challenge Time: 2:00pm - 2:30pm Session Chair: Paul Fussell Speaker(s): Kurt Stockinger, Mike Fisk, Thomas P. Caudell, Wes Bethel, Scott Campbell, Kesheng Wu, Stephen Lau, Paul Weber, Chris Davis, Rob Young, Rick Olinger, James E. Prewett Location: 611-614 Abstract: One of the most substantial challenges facing society today is information management and understanding. Our entry demonstrates the novel application of a set of diverse technologies aimed at improving our ability to more quickly and accurately identify and characterize malicious behavior hidden in large volumes of network flow data. The solution is a combination of data management, analysis and visual presentation technology. In our entry, we demonstrate: (1) scalable technology for performing data search/query operations; (2) use of machine-learning and neural-network techniques for feature identification and classification; (3) visual presentation paradigms well-suited to interactive exploration of high-dimensional data. Our entry is successfully used to perform queries, feature identification and characterization on 24 weeks of netflow data consisting of 1.1 billion records of 25 fields each. This paper can be found in the ACM and IEEE Digital Libaries Click here for ACM Click here for IEEE Chair/Speaker Details: Paul Fussell (Chair) Boeing Kurt Stockinger LBNL Mike Fisk LANL Thomas P. Caudell University of New Mexico Wes Bethel LBNL Scott Campbell LBNL/Nersc Kesheng Wu LBNL Stephen Lau LBNL/Nersc Paul Weber LANL Chris Davis LANL/UNM Rob Young UNM Rick Olinger UNM James E. Prewett UNM