Dr. Thai Le

Assistant Professor

207 Weir Hall
(662) 915-7438
thaile@olemiss.edu

Dr. Le received his Ph.D. in Information Sciences and Technology from the Pennsylvania State University in 2022. Before joining University of Mississippi, Dr. Le also worked at Alexa Privacy and was an ex-intern at Yahoo Research and VMWare OCTO. His research lies in data science and machine learning (ML). Dr. Le is especially interested in developing ML and AI models that are trustworthy to humans. Specific areas of interest include natural language processing, privacy and security for ML and explainable AI (XAI).

You can view Dr. Le’s homepage here.

Selected Publications

• SHIELD: Defending Textual Neural Networks against Multiple Black-Box Adversarial Attacks with Stochastic Multi-Expert Patcher
  Thai Le, Noseong Park, Dongwon Lee
  60th Annual Meeting of the Asso. for Comp. Linguistics (ACL), Dublin, Ireland, May 2022
  
• Perturbations in the Wild: Leveraging Human-Written Text Perturbations for Realistic Adversarial Attack and Defense
  Thai Le, Jooyoung Lee, Kevin Yen, Yifan Hu, Dongwon Lee
  Findings of 60th Annual Meeting of the Asso. for Comp. Linguistics (ACL-Findings), Dublin, Ireland, May 2022
  
• CAPS: Comprehensible Abstract Policy Summaries for Explaining Reinforcement Learning Agents
  Joe McCalmon, Thai Le, Sarra Alqahtani, Dongwon Lee
  Int’l Conf. on Autonomous Agents and Multiagent Systems (AAMAS), Virtual Event, May 2022
  Acceptance Rate: 26.4% (166/629)
• Socialbots on Fire: Modeling Adversarial Behaviors of Socialbots via Multi-Agent Hierarchical Reinforcement Learning
  Thai Le, Long Tran-Thanh, Dongwon Lee
  The ACM Web Conference (WWW), Virtual Event, April 2022
  Acceptance Rate: 17.7% (323/1,822)
• TuringBench: A Benchmark Environment for Turing Test in the Age of Neural Text Generation
  Adaku Uchendu, Zeyu Ma, Thai Le, Rui Zhang, Dongwon Lee
  Findings of Conf. on Empirical Methods in Natural Language Processing (EMNLP-Findings), Virtual Event, November 2021
  
• A Sweet Rabbit Hole by DARCY: Using Honeypots to Detect Universal Trigger’s Adversarial Attacks
  Thai Le, Noseong Park, Dongwon Lee
  59th Annual Meeting of the Asso. for Comp. Linguistics (ACL), Virtual Event, October 2021 (Oral)
  Acceptance Rate: 21.3% (713/3350)
• Authorship Attribution for Neural Text Generation
  Adaku Uchendu, Thai Le, Kai Shu, Dongwon Lee
  Conf. on Empirical Methods in Natural Language Processing (EMNLP), Virtual Event, November 2020
  Acceptance Rate: 22.4% (754/3,359)
• MALCOM: Generating Malicious Comments to Attack Neural Fake News Detection Models
  Thai Le, Suhang Wang, Dongwon Lee
  20th IEEE Int’l Conf. on Data Mining (ICDM), Virtual Event, November 2020
  Acceptance Rate: 9.8% (91/930)
• GRACE: Generating Concise and Informative Contrastive Sample to Explain Neural Network Model’s Prediction
  Thai Le, Suhang Wang, Dongwon Lee
  26th ACM SIGKDD Int’l Conf. on Knowledge Discovery and Data Mining (KDD), Virtual Event, August 2020
  Acceptance Rate: 16.9% (216/1,279)
• “Fake News” is Not Simply False Information: A Concept Explication and Taxonomy of Online Content
  Maria Molina, S. Shyam Soundar, Thai Le, Dongwon Lee
  American Behavioral Scientist, page 1-33, October 2019