Dongxu Zhang

Amherst, MA 01002

I am a research scientist at ASAPP . Previously I received my PhD in Computer Science at Umass Amherst , working with Professor Andrew McCallum. My thesis topic is to represent graphs in a compact, geometric-based embedding space. I am broadly interested in natural language processing and machine learning.

During my research career, I have had the great pleasure collaborating with Michael Boratko (UMass), Lorraine Li (UMass), Sunil Mohan (CZI), Subhabrata Mukherjee (now at Microsoft), Luna Dong (Amazon, now at Meta), Sujian Li (Peking University), and Dong Wang (Tsinghua University).

News

Jan 8, 2025	I am thrilled to announce that I have joined Optum (UnitedHealth Group) as a Lead Machine Learning Scientist. I am excited about the opportunity to contribute to the healthcare domain—an area that holds immense significance for all of us.
May 15, 2024	A paper about hallucination detection has been accepted by ACL 2024 findings.
Mar 27, 2023	Joined ASAPP.
Feb 28, 2023	Graduated from PhD!
Sep 14, 2022	Our paper has been accepted by Neurips 2022.

Selected Publications

NeurIPS

Modeling Transitivity and Cyclicity in Directed Graphs via Binary Code Box Embeddings

Dongxu Zhang, Michael Boratko, Cameron Musco, and Andrew McCallum

Advances in Neural Information Processing Systems 2022

Bib

  title = {Modeling Transitivity and Cyclicity in Directed Graphs via Binary Code Box Embeddings},
  author = {Zhang, Dongxu and Boratko, Michael and Musco, Cameron and McCallum, Andrew},
  journal = {Advances in Neural Information Processing Systems},
  year = {2022}
}

NeurIPS

Capacity and Bias of Learned Geometric Embeddings for Directed Graphs

Michael Boratko*, Dongxu Zhang*, Nicholas Monath, Luke Vilnis, Kenneth L. Clarkson, and Andrew McCallum

Advances in Neural Information Processing Systems 2021

Paper Bib Code

@article{boratko2021capacity,
  title = {Capacity and Bias of Learned Geometric Embeddings for Directed Graphs},
  author = {Boratko*, Michael and Zhang*, Dongxu and Monath, Nicholas and Vilnis, Luke and Clarkson, Kenneth L. and McCallum, Andrew},
  journal = {Advances in Neural Information Processing Systems},
  volume = {34},
  pages = {16423--16436},
  year = {2021},
}

NAACL

OpenKI: Integrating Open Information Extraction and Knowledge Bases with Relation Inference

Dongxu Zhang, Subhabrata Mukherjee, Colin Lockard, Xin Luna Dong, and Andrew McCallum

In Proceedings of NAACL-HLT 2019

Paper Bib Data

@inproceedings{zhang2019openki,
  title = {OpenKI: Integrating Open Information Extraction and Knowledge Bases with Relation Inference},
  author = {Zhang, Dongxu and Mukherjee, Subhabrata and Lockard, Colin and Dong, Xin Luna and McCallum, Andrew},
  data = {https://github.com/zhangdongxu/relation-inference-naacl19},
  booktitle = {Proceedings of NAACL-HLT},
  pages = {762--772},
  year = {2019},
  paper = {https://aclanthology.org/N19-1083/}
}

COLING

Bitext name tagging for cross-lingual entity annotation projection

Dongxu Zhang, Boliang Zhang, Xiaoman Pan, Xiaocheng Feng, Heng Ji, and Weiran Xu

In Proceedings of the 26th International Conference on Computational Linguistics: Technical Papers 2016

Paper Bib

@inproceedings{zhang2016bitext,
  title = {Bitext name tagging for cross-lingual entity annotation projection},
  author = {Zhang, Dongxu and Zhang, Boliang and Pan, Xiaoman and Feng, Xiaocheng and Ji, Heng and Xu, Weiran},
  booktitle = {Proceedings of the 26th International Conference on Computational Linguistics: Technical Papers},
  pages = {461--470},
  year = {2016},
  paper = {https://aclanthology.org/C16-1045/}
}