Shang Ma

I am Shang Ma, a PhD student in Computer Science and Engineering at the University of Notre Dame, advised by Prof. Yanfang (Fanny) Ye. I collaborate with Prof. Xusheng Xiao and my lab mates.

My research focuses on building structured AI to understand and defend against cyber fraud. I ground AI in graphs, code semantics, and human psychology to analyze digital threats and support safer interactions for people and AI agents. My work spans two connected directions:

App and AI Agent Security: Ad Fraud Detection (NDSS ’25), UI Agents for Software Testing (arXiv, AACL ’26), UI Agent Security (TAISAP).
Online Scam Understanding and Defense: Psychological Understanding (EMNLP’25, arXiv), Real-Time Detection (COLM ’26), Timely Intervention (arXiv).

See notes and tools we built for scams: Scam vs Fraud, Scam Daily, US Scam Map → see all

news

Sep 2026Agent+P accepted to the AACL 2026 Main Conference as an oral presentation.
Jul 2026PreScam accepted to COLM 2026.
Jun 2026Started a research internship in the Integrated Systems department at NEC Laboratories America, Princeton, working with Biplob Debnath.
Jan 2026Started a research internship in the Data Science & System Security department at NEC Laboratories America, Princeton, working with Wei Cheng.
Dec 2025Selected for the Meta Bug Bounty: Abuse Track.
Aug 2025PsyScam accepted to EMNLP 2025.
Oct 2024Started a research visit to Curiosity Lab, National University of Singapore, working with Zhenkai Liang.
Sep 2024Our work on detecting malware promotion via app promotion graphs accepted to NDSS 2025.
Mar 2024Started a research visit to ASE Lab, Peking University, working with Tao Xie.
Aug 2022Began my PhD journey.

experience

Research Intern · NEC Laboratories America

Integrated Systems
June–August 2026 · Princeton, New Jersey

Memory for enterprise proactive agent

Research Intern · NEC Laboratories America

Data Science & System Security
January–April 2026 · Princeton, New Jersey

AI-generated misinformation detection
Enterprise-level repository code question answering

selected publications

COLM

PreScam: transforming raw scam reports into structured conversations and scam kill chains

PreScam: A Benchmark for Predicting Scam Progression from Early Conversations

Weixiang Sun, Shang Ma, Yiyang Li, Tianyi Ma, Zehong Wang, Colby Nelson, Xusheng Xiao, and Yanfang Ye

In Conference on Language Modeling (COLM), 2026

arXiv Bib Code

@inproceedings{sun2026prescam,
  title = {PreScam: A Benchmark for Predicting Scam Progression from Early Conversations},
  author = {Sun, Weixiang and Ma, Shang and Li, Yiyang and Ma, Tianyi and Wang, Zehong and Nelson, Colby and Xiao, Xusheng and Ye, Yanfang},
  booktitle = {Conference on Language Modeling (COLM)},
  year = {2026},
}

AACL

Agent+P overview: node selector, UTG builder, symbolic plan generator, and UI explorer

Agent+P: Guiding UI Agents via Symbolic Planning

Shang Ma, Xusheng Xiao, and Yanfang Ye

In Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (AACL), Main Conference, 2026

Oral presentation

arXiv Bib Code

@inproceedings{ma2026agentp,
  title = {Agent+P: Guiding UI Agents via Symbolic Planning},
  author = {Ma, Shang and Xiao, Xusheng and Ye, Yanfang},
  booktitle = {Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (AACL), Main Conference},
  note = {Oral presentation},
  year = {2026},
}

EMNLP

PsyScam: psychological techniques illustrated in toll and job scams

PsyScam: A Benchmark for Psychological Techniques in Real-World Scams

Shang Ma, Tianyi Ma, Jiahao Liu, Wei Song, Zhenkai Liang, Xusheng Xiao, and Yanfang Ye

In Conference on Empirical Methods in Natural Language Processing (EMNLP), 2025

arXiv Bib Code

@inproceedings{ma2025psyscam,
  title = {PsyScam: A Benchmark for Psychological Techniques in Real-World Scams},
  author = {Ma, Shang and Ma, Tianyi and Liu, Jiahao and Song, Wei and Liang, Zhenkai and Xiao, Xusheng and Ye, Yanfang},
  booktitle = {Conference on Empirical Methods in Natural Language Processing (EMNLP)},
  year = {2025},
}

NDSS

AdGPE: app promotion graph, malware detection model, and promotion path inference

Careful About What App Promotion Ads Recommend! Detecting and Explaining Malware Promotion via App Promotion Graph

Shang Ma, Chaoran Chen, Shao Yang, Shifu Hou, Toby Jia-Jun Li, Xusheng Xiao, Tao Xie, and Yanfang Ye

In Network and Distributed System Security Symposium (NDSS), 2025

Bib PDF Code

@inproceedings{ma2025adgpe,
  title = {Careful About What App Promotion Ads Recommend! Detecting and Explaining Malware Promotion via App Promotion Graph},
  author = {Ma, Shang and Chen, Chaoran and Yang, Shao and Hou, Shifu and Li, Toby Jia-Jun and Xiao, Xusheng and Xie, Tao and Ye, Yanfang},
  booktitle = {Network and Distributed System Security Symposium (NDSS)},
  year = {2025},
}

Recent Preprints

Preprint

Scam scenario detection overview: incident report pretraining, transfer learning for conversations, and early-alert inference

The Anatomy of Scam Scenarios: Large-Scale Characterization and Conversation-Aware Detection

Shang Ma, Chen Yanai, Avichai Ben, Zichen Liu, Yanfang Ye, and Xusheng Xiao

arXiv preprint, 2026

arXiv Bib

@article{ma2026scenarios,
  title = {The Anatomy of Scam Scenarios: Large-Scale Characterization and Conversation-Aware Detection},
  author = {Ma, Shang and Yanai, Chen and Ben, Avichai and Liu, Zichen and Ye, Yanfang and Xiao, Xusheng},
  journal = {arXiv preprint},
  year = {2026},
  recent_preprint = {true},
}

Preprint

From UI to Code: Mobile Ads Detection via LLM-Unified Static-Dynamic Analysis

Shang Ma, Wei Cheng, Yanfang Ye, and Xusheng Xiao

arXiv preprint, 2026

arXiv Bib

@article{ma2026adwise,
  title = {From UI to Code: Mobile Ads Detection via LLM-Unified Static-Dynamic Analysis},
  author = {Ma, Shang and Cheng, Wei and Ye, Yanfang and Xiao, Xusheng},
  journal = {arXiv preprint},
  year = {2026},
  recent_preprint = {true},
}

I completed my undergraduate studies at Wuyuzhang Honors College, Sichuan University. Outside research, I enjoy road trips, going to the gym, and exploring history and military geography.

Google Scholar · LinkedIn · Email