Developing Adversary-Aware Classifiers for Android Malware Detection. Smartphones have become increasingly ubiquitous in people’s everyday life. However, it was reported that one in every five Android applications were actually malware, considering that Android has taken 88% market share of mobile phones. As an effective technique, machine learning has been widely adopted to detect Android malware. However, recent work suggests that deliberately-crafted malware makes machine learning ineffective ....Developing Adversary-Aware Classifiers for Android Malware Detection. Smartphones have become increasingly ubiquitous in people’s everyday life. However, it was reported that one in every five Android applications were actually malware, considering that Android has taken 88% market share of mobile phones. As an effective technique, machine learning has been widely adopted to detect Android malware. However, recent work suggests that deliberately-crafted malware makes machine learning ineffective. In this project, we propose to develop a series of new techniques, such as 1) Android contextual analysis, 2) wrapper-based hill climbing algorithm, and 3) ensemble learning, to solve this problem. The outcomes will help Australia gain cutting edge technologies in adversarial machine learning and mobile security.Read moreRead less
Defending AI based FinTech Systems against Model Extraction Attacks. This project aims to develop new methods for defending artificial intelligence (AI) based FinTech systems from highly potent and insidious model extraction attacks whereby an adversary can steal the AI model from the system to cause intellectual property (IP) violation, business advantage disruption, and financial loss. This can be achieved by examining various attack models, creating active and utility-preserving defences, and ....Defending AI based FinTech Systems against Model Extraction Attacks. This project aims to develop new methods for defending artificial intelligence (AI) based FinTech systems from highly potent and insidious model extraction attacks whereby an adversary can steal the AI model from the system to cause intellectual property (IP) violation, business advantage disruption, and financial loss. This can be achieved by examining various attack models, creating active and utility-preserving defences, and inventing non-removable watermarks on AI models. The outcomes are new tools for securing AI-based FinTech systems before deployment and tools for IP violation forensics post-deployment. Such capabilities are beneficial by improving the security and safety of FinTech systems and other nationally critical AI systems.Read moreRead less
Detecting Firmware Vulnerabilities in Smart Home Devices. 83% of Australians have smart home devices. 47% claim they have three or more. These devices are easily targeted by cyber-attacks, and searching for their vulnerabilities has become more crucial than ever. Our industry partner GPG is actively looking for ways to detect vulnerabilities in their smart home products, but have not found any existing methods that satisfy three critical requirements: 1) massive search, 2) cross platform detecti ....Detecting Firmware Vulnerabilities in Smart Home Devices. 83% of Australians have smart home devices. 47% claim they have three or more. These devices are easily targeted by cyber-attacks, and searching for their vulnerabilities has become more crucial than ever. Our industry partner GPG is actively looking for ways to detect vulnerabilities in their smart home products, but have not found any existing methods that satisfy three critical requirements: 1) massive search, 2) cross platform detection, and 3) finding unseen vulnerabilities. We therefore propose to use a series of new techniques such as efficient in-memory fuzzing, conditional formulas, and transfer learning to solve the above challenges. The project outcomes will help Australia gain cutting edge techniques in vulnerability detection. Read moreRead less
Developing an effective defence to cyber-reputation manipulation attacks. This project will develop new technologies for businesses to accurately identify fake internet reviews. Fake reviews, paid for and/or written with malicious intent, can cause irreparable damage to businesses resulting in revenue loss, consumer dissatisfaction or even closure of businesses. However they are difficult to identify, as they continuously evolve to avoid detection and the volume of Internet reviews makes analysi ....Developing an effective defence to cyber-reputation manipulation attacks. This project will develop new technologies for businesses to accurately identify fake internet reviews. Fake reviews, paid for and/or written with malicious intent, can cause irreparable damage to businesses resulting in revenue loss, consumer dissatisfaction or even closure of businesses. However they are difficult to identify, as they continuously evolve to avoid detection and the volume of Internet reviews makes analysis a monumental task. This project will provide advanced tools to detect fake website reviews and a cybersecurity system prototype ready to be used by industry, making Australia a leader in this field and resulting in a safer internet environment for all.Read moreRead less
Targeted Graph Embedding for Anomaly Detection in Large-scale Networks. This project aims to tackle the challenging problem of anomaly detection in large-scale networks by leveraging graph embedding techniques. It expects to deliver a series of innovative graph embedding algorithms targeting optimised anomaly detection. By addressing under-developed research challenges, such as the versatile types of anomalies and lack of anomaly labels, the established theories and devised methodologies will ad ....Targeted Graph Embedding for Anomaly Detection in Large-scale Networks. This project aims to tackle the challenging problem of anomaly detection in large-scale networks by leveraging graph embedding techniques. It expects to deliver a series of innovative graph embedding algorithms targeting optimised anomaly detection. By addressing under-developed research challenges, such as the versatile types of anomalies and lack of anomaly labels, the established theories and devised methodologies will advance frontier technologies in both graph anomaly detection and graph representation learning. By uncovering anomalies with high efficiency and accuracy, this project will contribute to multiple real applications from fake review detection to financial fraud identification, bringing both social and economic benefits.Read moreRead less