About Me

I'm a Ph.D. candidate in the Institute for Network Sciences and Cyberspace at Tsinghua University, advised by Prof. Haixin Duan in the Network and Information Security Lab (NISL).

I received my Bachelor degree in Commnication Engineering from Beijing University of Posts and Communications (BUPT) in 2018.

Currently, my research studies network security, with a particular interest in data-driven security and AI security.

Reserch Interests

Network Security; Data Driven Security; AI Security

Contact Me

• FIT 4-206, Tsinghua University, Beijing 100084, P.R.China

• liumx [AT] mails [DOT] tsinghua.ehu.cn

Other Links: Github | NISL

Publications

Conference

• Exploring the Characteristics and Security Risks of Emerging Emoji Domain Names, [ESORICS'22]
  Mingxuan Liu, Yiming Zhang, Baojun Liu, Haixin Duan
  

• Encrypted Malware Traffic Detection via Graph-based Network Analysis, [RAID'22]
  Zhuoqun Fu, Mingxuan Liu, Yue Qin, Jia Zhang, Yuan Zou, Qilei Yin, Qi Li, Haixin Duan
  (Pre-print Version)

• Ethics in Security Research: Visions, Reality, and Paths Forward, [EthiCS'22]
  Yiming Zhang, Mingxuan Liu, Mingming Zhang, Chaoyi Lu, Haixin Duan

• ValCAT: Generating Variable-Length Contextualized Adversarial Transformations using Encoder-Decode, [NAACL'22]
  Chuyun Deng, Mingxuan Liu, Yue Qin, Jia Zhang, Haixin Duan, Donghong Sun

• Detecting and Characterizing SMS Spearphising Attacks, [ACSAC'21]
  Mingxuan Liu, Yiming Zhang, Baojun Liu, Zhou Li, Haixin Duan, Donghong Sun
  (Acceptance rate: 24.5%)

• Argot: Generating Adversarial Readable Chinese Texts, [IJCAI'20]
  *Zihan Zhang, *Mingxuan Liu, Chao Zhang, Yiming Zhang, Zhou Li, Qi Li, Haixin Duan and Donghong Sun
  (* are co-first authors.)
  (Acceptance rate: 12.6%)

• Lies in the Air: Characterizing Fake-base-station Spam Ecosystem in China, [ACM CCS'20]
  Yiming Zhang, Baojun Liu, Chaoyi Lu, Zhou Li, Haixin Duan, Shuang Hao, Mingxuan Liu, Ying Liu, Dong Wang and Qiang Li
  Released Dataset
  (Acceptance rate: 16.9%)

• Casino Royale: A Deep Exploration of Illegal Online Gambling, [ACSAC'19]
  Hao Yang, Kun Du, Yubao Zhang, Shuang Hao, Zhou Li, Mingxuan Liu, Haining Wang, Haixin Duan, Yazhou Shi, Xiaodong Su, Guang Liu and Zhifeng Geng
  (Acceptance rate: 22.6%)

• TL;DR Hazard: A Comprehensive Study of Levelsquatting Scams, [SecureComm'19]
  Kun Du, Hao Yang, Zhou Li, Haixin Duan, Shuang Hao, Baojun Liu, Yuxiao Ye, Mingxuan Liu, Xiaodong Su, Guang Liu, Zhifeng Geng, Zaifeng Zhang and Jinjin Liang
  (Acceptance rate: 37.6%)

Jounal

• Automatic Generation of Adversarial Readable Chinese Texts, [TDSC'22]
  Mingxuan Liu, Zihan Zhang, Chao Zhang, Yiming Zhang, Zhou Li, Qi Li, Haixin Duan and Donghong Sun

Projects (in English)

• Automatic Adversarial Generation Project, [Huawei co.]
  Research on attack and defese technologies on adversarial texts

• Detection of Underground Website based on Natrual Langurage Processing (NLP), [Tsinghua]
  Completion of the website detection section based on semantic similarity comparison;
  This detection system has been deployed in Tsinghua Campus Network and Qi’anxin, and has been running continuously for more than 500 days, detecting thousands of blackmail pages per day

• Detection of Malicious Traffic based on Machine Learning Technology, [Huawei co.]
  Completion of detection based on features extracted from HTTP traffic;
  Resulted in a patent and a detection model that can be run at the gateway

• Detection of Malicious Encrypted Traffic, [Qi’anxin co.]
  Completion of detection system based on TLS flow features;
  Resulted in a patent (pending) and a detection system running at the gateway

Projects (in Chinese)

• 智能文本对抗项目, [华为]
  完成对抗文本生成的攻击和防御技术

• 基于自然语言处理的互联网黑灰产网站检测, [清华]
  完成基于语义相似度对比的页面检测部分；
  该检测系统已部署在清华校园网、奇安信，持续运行了 500 多天，每天可检出千量级的黑产页面

• 基于机器学习的网络恶意流量检测, [华为]
  完成基于特征提取的 HTTP 恶意流量检测部分；
  该部分已形成一篇专利和一个可以运行在现网中的检测模型

• 加密流量检测, [奇安信]
  完成基于 TLS 流特征的加密流量检测部分；
  该部分已提交一篇专利（审理中）和一个运行在现网中的检测模型

Services

External Reviewer

• NDSS'22, Euro S&P'22, Esorics'21, siMLA'21