Data-Intensive Distributed Systems LaboratoryCS Seminar
Date: March 11th, 2026
Time: 12:50pm
Room: SB111
Dr. Binghui Wang
Assistant Professor
Department of Computer Science
Illinois Institute of Technology
Talk Title
Trustworthy Graph Learning
Talk Abstract
Learning with graphs has attracted significant attention recently. Existing graph representation learning methods have achieved state-of-the-art performance on various graph-related tasks such as node/graph classification, link prediction, etc. However, recent studies show that graph representation learning methods are vulnerable to adversarial attacks.
In this talk, I will first introduce security attacks (e.g., evasion attacks, poisoning/backdoor attacks, etc.) and privacy attacks (e.g., private attribute inference, link inference, etc.) to graph representation learning. Then, I will talk about countermeasures against these attacks. In particular, I will first present certified defenses against evasion attacks and backdoor attacks. We prove the first (concurrent) certified robustness guarantee of any graph representation learning method against evasion attacks with graph structural perturbation. Our theoretical results are based on a recently proposed technique called randomized smoothing, which we extend to graph data. Moreover, we show that our certified robustness guarantee is tight.
Next, I will introduce our privacy-preserving graph representation learning framework. Our framework includes a primary learning task (e.g., node classification) and a privacy protection task (e.g., link protection), and our goal is to learn node representations such that they can be used to achieve high performance for the primary task, while obtaining performance for the privacy task close to random guessing. We formally formulate our goal via mutual information objectives, derive their tractable variational bounds, and train parameterized neural networks to obtain these bounds.
Speaker Bio
Binghui Wang is an Assistant Professor of Computer Science at the Illinois Institute of Technology since August 2021. His research lies at the intersection of security, privacy, and artificial intelligence, with a recent focus on provably secure and private AI and large language models (LLMs). His research has been published in premier security conferences including IEEE S&P, USENIX Security, ACM CCS, and NDSS, as well as leading AI and data science venues such as NeurIPS, ICML, ICLR, AAAI, CVPR, ECCV, KDD, and WWW. He is a recipient of the NSF CAREER Award (2024), Cisco Research Award (2022), and Amazon Research Award (2020). He has also been recognized with the Dean's Excellence in Research (2024) and named among the Global Top 50 Chinese Rising Stars in AI + X by Baidu Scholar (2022). His research has earned multiple Best Paper Awards (CCS'24, CVPRW'20), Honorable Mention (NDSS'19), and Oral presentations (AAAI'25, ICLR'24, CVPR'22).