"Innovative Security Framework Wins Grad Student Best Paper Award"

The best paper award at the 2022 IEEE International Conference on Trust, Privacy, and Security in Intelligent Systems and Applications went to a Machine Learning (ML) framework that detects security flaws without the computational overhead of traditional models. As scientists seek methods to uncover security flaws in software platforms automatically, they have used many of the same ML tools proven to be effective in other applications. Deep learning models known as transformers have significantly changed computer vision and Natural Language Processing (NLP). Significant breakthroughs in robotics and machine translation have been made because of Long Short-Term Memory (LSTM) neural networks. However, according to the researchers, these conventional methods for detecting security exploits require too much computing overhead for real-time security systems. In order to lessen this overhead, the team led by Princeton University developed a novel framework for exploit detection that involves public computer security databases and a mix of pattern-based techniques. The ML-FEED model has been proven to be 70 times faster than lightweight LSTMs and more than 75,000 times faster than transformers in the team's exploit detection tests. This article continues to discuss the ML-FEED model that won the best paper award at the 2022 IEEE International Conference on Trust, Privacy, and Security in Intelligent Systems and Applications. 

Princeton University reports "Innovative Security Framework Wins Grad Student Best Paper Award"