Rate My Professor Qiongkai Xu

Dr. Qiongkai Xu is a Lecturer in Natural Language Processing in the School of Computing, Faculty of Science and Engineering at Macquarie University. He earned his PhD in Natural Language Processing from the Australian National University, with a focus on Privacy Protection in Conversations. Previously, he served as a Research Fellow at the University of Melbourne and currently holds an Honorary Fellow position there.

His research focuses on Natural Language Processing, Privacy and Security, Machine Learning, and Data Mining. He specializes in auditing machine learning models by identifying and addressing privacy and security issues in ML and NLP models and their applications, as well as developing comprehensive evaluation theory and methods for these models. Specific interests include text watermarking and fingerprinting, backdoor and adversarial attacks, data leakage in language models, and performance evaluation. Xu maintains an active publication record in top conferences and journals such as ACL, EMNLP, NeurIPS, ICLR, AAAI, WWW, WSDM, and CIKM. Key publications include ALGEN: Few-shot inversion attacks on Textual Embeddings via Cross-Model Alignment and Generation (ACL 2025), Cut the deadwood out: backdoor purification via guided module substitution (EMNLP 2025), GRADA: graph-based reranking against adversarial documents attack (EMNLP 2025), HEAL: healthcare emergency assistants leveraging large language models (WWW Companion 2025), IDT: dual-task adversarial rewriting for attribute anonymisation (Computational Linguistics 2025), Generative Models are Self-Watermarked: Declaring Model Authentication through Re-Generation (TMLR 2024), Here’s a Free Lunch: Sanitizing Backdoored Models with Model Merge (Findings of ACL 2024), Security Challenges in Natural Language Processing Models (EMNLP Tutorial Abstracts 2023), Humanly Certifying Superhuman Classifiers (ICLR 2023), CATER: Intellectual Property Protection on Text Generation APIs via Conditional Watermarks (NeurIPS 2022), and Student Surpasses Teacher: Imitation Attack for Black-Box NLP APIs (COLING 2022). He presented a tutorial on Security Challenges in Natural Language Processing Models at EMNLP 2023. Awards include first place in the ALTA shared task (2015), second place in the AusDM shared task (2015), DAAD scholarship (2022), Data61 top-up scholarship (2017), and Research Pitching Session Winner at Macquarie University (2022 and 2024). He is Primary Chief Investigator on projects Legal Compliant and Secure Smart Contract Generation with LLMs (2025-2029) and Climate Litigation Risk: AI-Enhanced Greenwashing Detection.