Sundong Kim (김선동)

About Me

I am a Young Scientist Fellow at Data Science Group, Institute for Basic Science, South Korea. Over the past few years, I worked on representation learning, predictive analytics, and democratizing AI for social goods (e.g., Customs). In these days, I am enthusiastic about understanding and building a human-like AI with abstraction and reasoning. Prior to IBS, I obtained my Ph.D. from KAIST in 2019.

E-mail: sundong (at)
Office: B233, Data Science Group, Institute for Basic Science (IBS)
Tel: (+82)-042-878-9202

I will join Graduate School of AI, GIST as an assistant professor (Nov 2022).

Prospective students: see this page.

[CV] [Google Scholar]

Selected Publications (Full list)

Fraud Detection

  • Knowledge Sharing via Domain Adaptation in Customs Fraud Detection
    Sungwon Park, Sundong Kim, Meeyoung Cha
    AAAI 2022

  • Customs Fraud Detection in the Presence of Concept Drift
    Tung-Duong Mai, Kien Hoang, Aitolkyn Baigutanova, Gaukhartas Alina, Sundong Kim
    ICDM 2021 (IncrLearn Workshop)
    [Link] [Video]

  • Active Learning for Human-in-the-loop Customs Inspection
    Sundong Kim, Tung-Duong Mai, Sungwon Han, Sungwon Park, Thi Nguyen Duc Khanh, Jaechan So, Karandeep Singh, Meeyoung Cha
    IEEE Transactions on Data and Knowledge Engineering, 2022. (SCI, IF=6.977)
    [Link] [Github]

  • DATE: Dual Attentive Tree-aware Embedding for Customs Fraud Detection
    Sundong Kim*, Yu-Che Tsai*, Karandeep Singh, Yeonsoo Choi, Etim Ibok, Cheng-Te Li, Meeyoung Cha
    KDD 2020 (Applied Data Science)
    [Link] [PDF] [Slides] [Talk] [Github] [Project] [Promotional video] [WCO News] [Press]

User Modeling

  • Revisit Prediction by Deep Survival Analysis
    Sundong Kim, Hwanjun Song, Sejin Kim, Beomyoung Kim, Jae-Gil Lee
    PAKDD 2020
    [Link] [PDF] [Slides] [Talk] [Github]

  • Utilizing In-Store Sensors for Revisit Prediction
    Sundong Kim, Jae-Gil Lee
    ICDM 2018
    (Selected as one of the best papers in ICDM 2018)
    (Longer version: Knowledge and Information Systems 2020 [Link])
    [Link] [PDF] [Slides] [Poster] [Talk] [Github]

  • Friend Recommendation with a Target User in Social Networking Services
    Sundong Kim
    ICDE 2015 (Ph.D. Symposium)
    [Link] [PDF] [Slides]

Embedding Learning

  • FedX: Unsupervised Federated Learning with Cross Knowledge Distillation
    Sungwon Han, Sungwon Park, Fangzhao Wu, Sundong Kim, Chuhan Wu, Xing Xie, Meeyoung Cha
    ECCV 2022
    [PDF] [Github] [Poster] [Video]

  • Embedding Heterogeneous Hierarchical Structure
    Sundong Kim
    IC2S2 2021 (Extended Abstract)
    [PDF] [Github]

  • Improving Unsupervised Image Clustering With Robust Learning
    Sungwon Park*, Sungwon Han*, Sundong Kim, Danu Kim, Sungkyu Park, Seunghoon Hong, Meeyoung Cha
    CVPR 2021
    [Link] [Github]

  • Mitigating Embedding and Class Assignment Mismatch in Unsupervised Image Classification
    Sungwon Han, Sungwon Park, Sungkyu Park, Sundong Kim, Meeyoung Cha
    ECCV 2020
    [PDF] [Supplementary] [Github] [Talk] [Summary]

  • Neural User Embedding From Browsing Events
    Mingxiao An, Sundong Kim
    ECML-PKDD 2020 (Applied Data Science)
    [PDF] [Talk]

XAI & Data Science

  • Explainable Product Classification for Customs
    Eunji Lee, Sundong Kim, Sihyun Kim, Sungwon Park, Meeyoung Cha, Soyeon Jung, Suyoung Yang, Yeonsoo Choi, Sungdae Ji, Minsoo Song, Heeja Kim
    Korean Artificial Intelligence Association, 2021
    [Link] [Demo]

  • Disruption in the Chinese E-Commerce During COVID-19
    Yuan Yuan, Muzhi Guan, Zhilun Zhou, Sundong Kim, Meeyoung Cha, Depeng Jin, Yong Li
    Frontiers in Computer Science, 2021

  • Response to COVID-19 with Probabilistic Programming
    Assem Zhunis, Tung-Duong Mai, Sundong Kim
    IC2S2 2021 (Extended Abstract)
    Long version: arXiv:2106.00192
    [Link] [Abstract] [Github]

Active Learning

  • Coherence-based Label Propagation over Time Series for Accelerated Active Learning
    Yooju Shin, Susik Yoon, Sundong Kim, Hwanjun Song, Jae-Gil Lee, Byung Suk Lee
    ICLR 2022

  • Carpe Diem, Seize the Samples Uncertain “at the Moment” for Adaptive Batch Selection
    Hwanjun Song, Minseok Kim, Sundong Kim, Jae-Gil Lee
    CIKM 2020
    [Link] [PDF] [Talk] [Github] [Press]

  • Ada-Boundary: Accelerating the DNN Training via Adaptive Boundary Batch Selection
    Hwanjun Song, Sundong Kim, Minseok Kim, Jae-Gil Lee
    Machine Learning, 2020
    [Link] [Slides] [Talk] [Github]

Academic Services

Talks and Videos

  • Research:
    • A Glimpse of the Happy ARC Day 🧩: [Video]
    • Developing artificial intelligence to support human intelligence: [PDF]
    • Dual attentive tree-aware embedding for customs fraud detection: [KACT Webinar] [English] [French]
    • Revisit prediction by deep survival analysis [Talk]
    • Utilizing in-store sensors for revisit prediction [Talk]
  • Others:
    • Why did I become a researcher? [Talk] [PDF]
    • What I have done so far, what I am interested in. [PDF]
    • Machine learning approach for customs fraud detection [PDF]
    • Tools and approaches for applied science in the era of big data [PDF]


I have been fortunate to work with many gifted students:

Graduate Students at IBS/KAIST Data Science Group (DS Group):

  • Sungwon Park (→ PhD at DS Group, Intern at MSRA), 2020-2022
  • Sungwon Han (→ PhD at DS Group, Intern at MSRA, UToronto), 2020-2022
  • Eunji Lee (→ Visiting UCLA), 2021-2022

Undergraduate Students at IBS/KAIST Data Science Group:

Graduate Students at KAIST Data Mining Group:

  • Yooju Shin (→ PhD Student at KAIST DM Lab), 2017, 2021-2022
  • Minseok Kim (→ Applied Scientist at Amazon Alexa AI), 2017, 2019-2020
  • Hwanjun Song (→ Research Scientist at Naver AI Lab), 2019-2022

Undergraduate Students at KAIST Data Mining Group: