Memgraph logo
Webinar

Graph-based vulnerability discovery

About the event
Graph-based machine learning is increasingly adopted for automatic security vulnerability discovery due to its ability to represent programs as graphs that preserve semantic information while eliminating unnecessary syntactic elements. Graph neural networks (GNNs) exploit these graph structures for effective representation learning, creating a powerful combination. However, challenges arise, such as the need for large datasets and the rapid growth of program graph sizes.

In this community call, Dr. Tom Ganz will discuss leveraging Memgraph to generate vast amounts of data from an engineering perspective and modeling the problem from a research perspective to manage large graphs. Specifically, he'll show how to transform current inaccurate graph-based static analyzers into context-sensitive, patch-based static analyzers, using Memgraph as the backbone for data mining.

About the speaker
Tom Ganz is a security engineer at Amazon, with a strong theoretical background in machine learning and a particular focus on static program analysis. He holds a Ph.D. in Software Security and Machine Learning from TU Berlin and has broad expertise in various fields of computer science. Tom has worked as a software developer, security engineer, and data scientist at several renowned companies before.
© 2024 Memgraph Ltd. All rights reserved.