Dongjun Lee

2066, Seobu-ro, Jangan-gu

Suwon-si, Gyeonggi-do, Republic of Korea

As a graduate researcher within the MainLab at Department of Electrical and Computer Engineering at Sungkyunkwan University (SKKU), I was advised under Prof. Jaekwang Kim. My scholarly journey commenced in Mar. 2016 and culminated in Feb. 2022 with the attainment of a B.S. degree from Daegu Catholic University, South Korea. Throughout this period, my focus centered on the exploration and application of deep learning techniques to the processing and analysis of signal processing modalities, notably electroencephalography (EEG) signals.

My research interest is directed towards four primary fields: (1) recommendation systems, (2) generative models, (3) self-supervised learning, and (4) large language models. The crux of my work is the enhancement of user personalization and experience, driven by the ethos of "Making the World a Better Place." As I seek to achieve this goal, I am in pursuit of an opportunity with a forward-thinking groups or team where I can apply my expertise in Machine Learning, and Deep Learning to contribute to our collective future. Presently, I am employed at MaumAI, serving as an artificial intelligence scientist, where my role involves orchestrating language models tailored for distinct business objectives. Should you wish to engage in collaboration or have inquiries, I welcome you to reach out.

news

Mar 1, 2024	“Retrieval-Based Disease Prediction for Myocardial Injury after Noncardiac Surgery: Leveraging Language Models as Diagnostic Tools” was accepted at the AAAI 2024 Spring Symposium on Clinical Foundation Models in Stanford University, Stanford, California (Poster)!
Dec 7, 2023	“Self-Interactive Attention Netwroks via Factorization Machinces for Click-Through Rate Prediction” was presented at the ISIS 2023 international conference (Oral)! and won the best paper award.
Nov 27, 2023	A patent “Elevating CTR Prediction: Field Interaction, Global Context Integration, and High-Order Representations” have been accepted in ACM/SIGAPP Symposium On Applied Computing (SAC) !
Nov 3, 2023	A patent “순차적 추천을 위한 계층적 대조 학습 방법 및 계층적 대조 학습 장치” have been filed in Korea.
Oct 23, 2023	“AmpliBias:Mitigating Dataset Bias through Bias Amplification in Few-Shot Learning for Generative Models” was presented at the CIKM 2023 international conference in Birmingham (Poster)!

selected publications

How Important is Periodic Model Update in Recommender System?

Hyunsung Lee, Sungwook Yoo, Dongjun Lee, and 1 more author

In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2023

Abs

In real-world recommender model deployments, the models are typically retrained and deployed repeatedly. It is the rule-of-thumb to periodically retrain recommender models to capture up-to-date user behavior and item trends. However, the harm caused by delayed model updates has not been investigated extensively yet. in this perspective paper, we formulate the delayed model update problem and quantitatively demonstrate the delayed model update actually harms the model performance by increasing the number of cold users and cold items increase and decreasing overall model performances. These effects vary across different domains having different characteristics. Upon these findings, we further argue that although the delayed model update has negative effects on online recommender model deployment, yet it has not gathered enough attention from research communities. We argue our verification of the relationship between the model update cycle and model performance calls for further research such as faster model training, and more efficient data pipelines to keep the model more up-to-date with the latest user behaviors and item trends.
AmpliBias: Mitigating Dataset Bias through Bias Amplification in Few-Shot Learning for Generative Models

Donggeun Ko, Dongjun Lee, Namjun Park, and 3 more authors

In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, 2023

Abs

Deep learning models exhibit a dependency on peripheral attributes of input data, such as shapes and colors, leading the models to become biased towards these certain attributes that result in subsequent degradation of performance. In this paper, we alleviate this problem by presenting sysname, a novel framework that tackles dataset bias by leveraging generative models to amplify bias and facilitate the learning of debiased representations of the classifier. Our method involves three major steps. We initially train a biased classifier, denoted as f_b, on a biased dataset and extract the top-K biased-conflict samples. Next, we train a generator solely on a bias-conflict dataset comprised of these top-K samples, aiming to learn the distribution of bias-conflict samples. Finally, we re-train the classifier on the newly constructed debiased dataset, which combines the original and amplified data. This allows the biased classifier to competently learn debiased representation. Extensive experiments validate that our proposed method effectively debiases the biased classifier.