About me

Welcom to my homepage!

My name is Ming Xu. I am a research fellow in NUS (National University of Singapore), working with Prof. Jiaheng Zhang and Prof. Jin Song Dong. I received my Ph.D. Degree from Fudan University under the supervision of Prof. Weili Han.

I am working to protect the data safety of everyone online. My interests span security, privacy, fraud, misinformation, and practical applications of agentic AIs. Several works are listed below.

• Post-accidents (Digital Forensics): Post-incident data detection and anomaly tracing are essential components of modern digital forensics. As many sophisticated attacks evade preventive defenses and only surface through subtle log-level signals, systematic log analysis is crucial for reconstructing attack timelines, identifying root causes, and revealing multi-step intrusion chains. Key components include Agentic SIEM-rule creation ICSE’26, SIEM-rule conversion WWW’26, and LLM-based Web Anomaly Detectors ASE’24, which together enable adaptive, intelligent, and scalable threat provendance.

• Pre-accidents (Robust Authentications): Designing the next-generation robust authentication mechanisms by uncovering the behavioral and structural foundations of human-chosen passwords. Our work reveals the inherent patterns of textual passwords such as “p@ssw0rd4ever” CCS’21, exposes real-world password guessing behaviors and practical attack scenarios USENIX SEC’23, TDSC’25, and identifies emerging privacy risks within probabilistic strength meters (PSMs) AsiaCCS’25.

If you are interested in joining or remotely working with us, feel free to drop an email to me: [email protected]

Broader Research Interests

• Data-driven Security
• Agentic AI for Cyber Security
• Textual Password Security
• Cyber Threat Intelligence