About Me
My name is Sungyong Seo and I am a Software Engineer at Google Cloud AI. Prior to joining Google, I was a Ph.D. student in the Computer Science Department of University of Southern California under the supervision of Professor Yan Liu. I also spent time at the NYU Center for Data Science to collaborate with Professor Kyunghyun Cho.
I am interested in machine learning problems in general.
More specifically, my research mainly focuses on principled techniques for extracting knowledge from the complex structures or networks and using it to build predictive models – in that way merging insights both from data mining and machine learning. Furthermore, combining the structural knowledge with temporal behaviors is the topic which I am interested in. I am working on the following topics:
- Physics-inspired machine learning
- Spatiotemporal data mining
- Graph-based neural networks
- Recommendation systems
Education
Aug. 2015 - June 2021
University of Southern California
Computer Science Ph.D.
Aug. 2012 - Dec. 2014
University of Michigan, Ann Arbor
Electrical Engineering M.S.
Jan. 2008 - May 2008
Nanyang Technological University
Singapore TF-NTU LEaRN Program Visiting Student
Mar. 2005 - Feb. 2012
Seoul National University
Electrical Engineering B.S., Physics (Minor)
Publications
Conferences
- Controlling Neural Networks with Rule Representations
Sungyong Seo, Sercan O Arik, Jinsung Yoon, Xiang Zhang, Kihyuk Sohn, Tomas Pfister
Neural Information Processing Systems (NeurIPS) 2021.
[arXiv Paper] [Code] [K1st World Symposium]
- Physics-aware Spatiotemporal Modules with Auxiliary Tasks for Meta-Learning
Sungyong Seo, Chuizheng Meng, Sirisha Rambhatla, Yan Liu
International Joint Conference on Artificial Intelligence (IJCAI) 2021.
NeurIPS Workshop Machine Learning and the Physical Sciences (ML4PS) 2020.
[Paper] [Workshop Paper] [Workshop Poster]
- Physics-aware Difference Graph Networks for Sparsely-Observed Dynamics
Sungyong Seo*, Chuizheng Meng*, Yan Liu
International Conference on Learning Representations (ICLR) 2020.
[Paper] [OpenReview] [Code] [Bibtex]
- A Deep Structural Model for Analyzing Correlated Multivariate Time Series
Changwei Hu, Yifan Hu, Sungyong Seo
IEEE International Conference on Machine Learning and Applications (ICMLA) 2019.
[Paper] [Bibtex]
- Social Bots for Online Public Health Interventions
Ashok Deb, Anuja Majmundar, Sungyong Seo, Akira Matsui, Rajat Tandon, Shen Yan, Jon-Patrick Allem, Emilio Ferrara
IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) 2018.
[Paper] [Bibtex]
- Automatically Inferring Data Quality for Spatiotemporal Forecasting
Sungyong Seo, Arash Mohegh, George Ban-Weiss, Yan Liu
International Conference on Learning Representations (ICLR) 2018.
[Paper] [OpenReview] [Bibtex]
- Partially Generative Neural Networks for Gang Crime Classification with Partial Information
Sungyong Seo, Hau Chan, P. Jeffrey Brantingham, Jorja Leap, Phebe Vayanos, Milind Tambe, Yan Liu
AAAI/ACM Conference on AI, Ethics, and Society (AIES) 2018. (Oral presentation)
[Paper] [Bibtex]
- CSI: A Hybrid Deep Model for Fake News Detection
Sungyong Seo*, Natali Ruchansky*, Yan Liu
ACM International Conference on Information and Knowledge Management (CIKM) 2017.
[Paper] [Bibtex]
- Interpretable Convolutional Neural Networks with Dual Local and Global Attention for Review Rating Prediction
Sungyong Seo, Jing Huang, Hao Yang, Yan Liu
ACM Conference on Recommender Systems (RecSys) 2017.
[Paper] [Bibtex]
Workshops or Preprints
- Coronavirus on social media: Analyzing misinformation in Twitter conversations
Karishma Sharma, Sungyong Seo, Chuizheng Meng, Sirisha Rambhatla, Aastha Dua, Yan Liu
[arXiv Paper][Project Page][Fortune Eye on AI]
- Contextual Understanding of Homicide Reports in Los Angeles County
Sungyong Seo, Umang Gupta, Jiageng Zhu, P. Jeffrey Brantingham, Yan Liu
SoCal NLP Symposium 2019.
[Symposium page]
- Data-driven Temporal Attribution Discovery of Temperature Dynamics based on Attention Networks
Sungyong Seo, Jiachen Zhang, George Ban-Weiss, Yan Liu
International Workshop on Climate Informatics (CI) 2019.
[Workshop page]
- Differentiable Physics-informed Graph Networks
Sungyong Seo and Yan Liu
ICLR Workshop on Representation Learning on Graphs and Manifolds (ICLR-RLGM) 2019.
AAAI 2021 Spring Symposium on Combining Artificial Intelligence and Machine Learning with Physics Sciences (AAAI-MLPS) 2021.
[arXiv Paper] [Workshop page] [Symposium Proceedings] [Code]
- Data Quality Network for Spatiotemporal Forecasting
Sungyong Seo, Arash Mohegh, George Ban-Weiss, Yan Liu
NeurIPS Workshop on Deep Learning for Physical Sciences (NeurIPS-DLPS) 2017.
[Paper]
- Graph Convolutional Autoencoder with Recurrent Neural Networks for Spatiotemporal Forecasting
Sungyong Seo, Arash Mohegh, George Ban-Weiss, Yan Liu
International Workshop on Climate Informatics (CI) 2017.
[Paper]
- Representation Learning of Users and Items for Review Rating Prediction Using Attention-based Convolutional Neural Network
Sungyong Seo, Jing Huang, Hao Yang, Yan Liu
SDM Workshop on Machine Learning Methods for Recommender Systems (SDM-MLRec) 2017.
[Paper]