Hello I'm Mogan Gim who is currently a Assistant Professor in Hankuk University of Foreign Studies, Department of Biomedical Engineering (BME). I earned my Ph.D. in Computer Science from Korea University under the guidance of Professor Kang, who served as my principal investigator. During my Ph.D program under Professor Kang's supervision, I took part in various collaborative projects including SK pucore's SMARTUS development project and Sony AI's Gastronomy Project.
My title for my doctoral dissertation is Deep Learning Architectures using Set-Oriented Data Representations . As the title says, my research interests revolve around solving real-world problems using set-oriented data representations. I primarily tackle such problems by devising a deep learning-based approach that addresses their domain-specific aspects. I also focus on developing novel deep learning architectures that best utilizes those aspects for not only enhancing their task performance but also model explainability.
During my Ph.D program, I have engaged in several applied data science projects of diverse domains including Food Informatics, Drug Discovery, Material Science and so on. I enjoy learning not only the advancements of cutting-edge deep learning methodologies but also unexplored fields of knowledge. Though I recieved my Bachelor's Degree in Computer Science, I have seldom restrained myself from being intrigued in other fields of science such as organic chemistry, molecular biology, culinary studies, food nutrition as on. I believe it is fascinating to infuse such knowledge in developing AI-driven systems that can contribute to society. These academic characteristics are the key ingredients in my previous publications and current research proejcts.
Currently, my central research interest is Digital Health, a multi-dicsiplinary applied data science domain that spans Bioinformatics, Cheminformatics and Food Informatics. My utlimate goal is to pursue my research towards promoting social health through development of a AI-driven Drug Discovery Pipeleine and Precision Nutrition System.
[CV] [Email] [Scholar] [Github] [LinkedIn] [ORCID] [Office Location (Room 513)]
Research Area
Data Science and AI Applications
Sets, Graphs, Generative Models, LLMs
Digital Health
Drug Discovery
Precision Nutrition
Bioinformatics (major)
Food Informatics (major)
Cheminformatics (major)
Material Informatics (minor)
Finance Informatics (minor)
Law Informatics (minor)
Seungheun Baek*, Soyon Park*, Yan Ting Chok, Junhyun Lee, Jueon Park, Mogan Gim**, Jaewoo Kang**, "Cradle-VAE: Enhancing Single-Cell Gene Perturbation Modeling with Counterfactual Reasoning-based Artifact Disentanglement", arxiv, https://arxiv.org/abs/2409.05484
Mogan Gim*, Jueon Park*, Soyon Park, Sanghoon Lee, Seungheun Baek, Junhyun Lee, Ngoc-Quang Nguyen and Jaewoo Kang. "MolPLA: A Molecular Pretraining Framework for Learning Cores, R-Groups and their Linker Joints", ISMB 2024, July 12--16, 2024, Montreal, Canada (Full Paper Accepted, Oral Presentation, Bioinformatics, vol. 40, https://doi.org/10.1093/bioinformatics/btae256)
Mogan Gim, Junseok Choi, Seungheun Baek, Jueon Park, Chaeeun Lee, Minjae Ju, Sumin Lee and Jaewoo Kang. "ArkDTA: Attention Regularization guided by non-Covalent Interactions for Explainable Drug-Target Binding Affinity Prediction", ISMB/ECCB 2023, July 23--27, 2023, Lyon, France, (Full Paper Accepted, Oral Presentation, Bioinformatics, vol. 39, https://doi.org/10.1093/bioinformatics/btad207)
Mogan Gim*, Donghee Choi*, Kana Maruyama, Jihun Choi, Hajung Kim, Donghyeon Park and Jaewoo Kang. "RecipeMind: Guiding Ingredient Choices from Food Pairing to Recipe Completion using Cascaded Set Transformer", CIKM '22: Proceedings of the 31st ACM International Conference on Information and Knowledge Management, October 17--21, 2022, Atlanta, GA, USA, https://doi.org/10.1145/3511808.3557092 (Full Paper Accepted, Oral Presentation)
Mogan Gim*, Donghyeon Park*, Michael Spranger, Kana Maruyama and Jaewoo Kang. "RecipeBowl: A Cooking Recommender for Ingredients and Recipes using Set Transformer", IEEE Access, vol. 9, October 2021, https://doi.org/10.1109/ACCESS.2021.3120265'
Projects
Sony Research [2020 ~ 2024]
Collaborative Work with Sony AI Gastronomy Project
CookingSense: A Culinary Knowledgebase with Multidisciplinary Assertions (paper)
KitchenScale: Learning to predict ingredient quantities from recipe contexts (paper, github)
RecipeMind: Guiding Ingredient Choices from Food Pairing to Recipe Completion using Cascaded Set Transformer (paper, github)
RecipeBowl: A Cooking Recommend for Ingredients and Recipes Using Set Transformer (paper, github)
SK pucore [2018~ 2021]
Collaborative Work with SK pucore's SMARTUS Development Project (as Lead Resarcher/Developer)
Improved the SMARTUS System with Enhanced Prediction Model and Recomendation Algorithm
Developed and Deployed a Server-based System fo Urethane Recipe Formulation (SMARTUS)
Developed a Deep Learning Mdoel for Property Prediction of Polyurethane Products
Developed a Recommendation Algorithm for Polyurethane Recipe Formulations
SK Nexilis [2020~ 2021]
Collaborative Work with SK Nexilis' Development Project (as Developer)
Miscellaneous
Patents
METHOD AND SERVER FOR FOOD INGREDIENT PAIRING PREDICTION USING SIAMESE NEURAL NETWORk (10-2019-0158385), https://doi.org/10.8080/1020230049561
METHOD AND APPARATUS FOR BINDING AFFINITY PREDICTION BASED ON NON-COVALENT INTERACTION (10-2023-0049561), https://doi.org/10.8080/1020230049561
Competitions
[2018] NCI-CPTAC Multi-omics Enabled Sample Mislabeling Correction Challenge, PrecisionFDA
[2020] Preterm Birth Prediction: Transcriptomics, DREAM Challenge, 8th Place
[2022] Preterm Vaginal Microbiome Challenge, DREAM Challenge, 5th Place
Paper Review
[2024] BMC Bioinformatics, (IF: 3.169)
[2024] Information Processing and Management, (IF: 6.222)
Talk
Lecture
[2024-2] Digital Engineering
[2024-2] BME Electronic Circuits
Grants