Xin Liu

HI! I am a researcher at Supcon working on Industrial AI. Our team is building a suite of industrial intelligent agents based on HGT fostering steady and sustainable business growth, including:

HR Agent Data Agent Biding Agent Selling Agent Meeting Agent Purchasing Agent EfficiencyPi

My current work mainly focuses on using (hyper-)graph, ontology and large language models (LLMs) to encapsulate enterprise knowledge, business entities, and operational logic, enabling structured reasoning and task planning while preserving explainability and execution determinism. I am particularly interested in graph retrieval-augmented generation (GraphRAG) methods and graph foundation models (GFMs), bridging the gap between cutting-edge AI research and real-world industrial applications.

Before joining Supcon, I received my Ph.D. from Tsinghua University in 2025, advised by Prof. Wei Xu, where I worked on fraud detection, multi-party computation, graph algorithms, and graph neural networks.

selected publications

arXiv

Hypergraph Enterprise Agentic Reasoner over Heterogeneous Business Systems

Ling Wang^*, Xin Liu^*, Songnan Liu, and 8 more authors

arXiv preprint arXiv:2605.14259, 2026

arXiv Bib HTML

@article{wang2026hypergraphenterpriseagenticreasoner,
  title = {Hypergraph Enterprise Agentic Reasoner over Heterogeneous Business Systems},
  author = {Wang, Ling and Liu, Xin and Liu, Songnan and Wang, Jianan and Cheng, Cheng and Zhu, Yihan and Li, Enyu and Xiao, Yu and Xie, Jiangyong and Yan, Duogong and Chen, Jiangyi},
  journal = {arXiv preprint arXiv:2605.14259},
  year = {2026},
}

arXiv

Detecting LLM-Generated Spam Reviews by Integrating Language Model Embeddings and Graph Neural Network

Xin Liu^*, Rongwu Xu^*, Xinyi Jia, and 4 more authors

arXiv preprint arXiv:2510.01801, 2025

arXiv Bib

@article{liu2025detecting,
  title = {Detecting LLM-Generated Spam Reviews by Integrating Language Model Embeddings and Graph Neural Network},
  author = {Liu, Xin and Xu, Rongwu and Jia, Xinyi and Liao, Jason and Sun, Jiao and Huang, Ling and Xu, Wei},
  journal = {arXiv preprint arXiv:2510.01801},
  year = {2025},
}

CIKM

Collaborative Fraud Detection on Large Scale Graph Using Secure Multi-Party Computation

Xin Liu, Xiaoyu Fan, Rong Ma, and 4 more authors

In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, 2024

Bib PDF

@inproceedings{liu2024collaborative,
  title = {Collaborative Fraud Detection on Large Scale Graph Using Secure Multi-Party Computation},
  author = {Liu, Xin and Fan, Xiaoyu and Ma, Rong and Chen, Kun and Li, Yi and Wang, Guosai and Xu, Wei},
  booktitle = {Proceedings of the 33rd ACM International Conference on Information and Knowledge Management},
  pages = {1473--1482},
  year = {2024},
}

WWW

No place to hide: Catching fraudulent entities in tensors

Yikun Ban, Xin Liu, Ling Huang, and 3 more authors

In The World Wide Web Conference, 2019

Bib PDF Poster

@inproceedings{yikun2019no,
  title = {No place to hide: Catching fraudulent entities in tensors},
  author = {Ban, Yikun and Liu, Xin and Huang, Ling and Duan, Yitao and Liu, Xue and Xu, Wei},
  booktitle = {The World Wide Web Conference},
  pages = {83--93},
  year = {2019},
}