Hi!👋

I’m a third-year PhD student in Computer Science and Engineering at UC San Diego.

My research focuses on natural language processing and generative AI. I am particularly interested in leveraging Large Language Models (LLMs) for real-world applications and extracting knowledge from transformer and state space model based LLMs. I’m also keen on enhancing the textual capabilities of multimodal models, especially in text-to-image generation models.

Before UCSD, I worked as a software engineer on the Query Understanding team at Coupang (the largest e-commerce company in Korea), developing recommendation systems using large-scale user behavior data and product descriptions. I hold a B.S. and M.S. in Computer Science and Engineering from Sogang University.

[CV] [Google scholar]

Publications

Conference Proceedings

• SYMPTOMIFY: Transforming Symptom Annotations with Language Model Knowledge Harvesting. Bosung Kim and Ndapa Nakashole. 2023. (EMNLP Findings 2023) 📄💻

• Zero-shot Triplet Extraction by Template Infilling. Bosung Kim, Hayate Iso, Nikita Bhutani, Estevam Hruschka, Ndapa Nakashole, and Tom Mitchell. 2023. (IJCNLP-AACL 2023) 📄💻

• Chride at SemEval-2023 Task 10: Fine-tuned DebertaV3 on Detection of Online Sexism with Hierarchical Loss. Letian Peng and Bosung Kim. 2023. (ACL 2023 SemEval Workshop) 📄💻

• Data Augmentation for Rare Symptoms in Vaccine Side-Effect Detection. Bosung Kim and Ndapa Nakashole. 2022. (ACL 2022 BioNLP Workshop) 📄💻

• Query Reformulation for Descriptive Queries of Jargon Words Using a Knowledge Graph based on a Dictionary. Bosung Kim, Hyewon Choi, Haeun Yu, and Youngjoong Ko. 2021. (CIKM 2021) 📄

• Commonsense Knowledge Augmentation for Low-Resource Languages via Adversarial Learning. Bosung Kim*, Juae Kim*, Youngjoong Ko, and Jungyun Seo. 2020. (AAAI 2021) (*Equal contribution) 📄💻

• Multi-task learning for Knowledge Graph Completion using Pre-trained Language. Bosung Kim, Taesuk Hong, Youngjoong Ko, and Jungyun Seo. 2020. (COLING 2020) 📄💻

Journal Articles

• Novel Regularization Method for the Class Imbalance Problem. Bosung Kim, Youngjoong Ko, and Jungyun Seo. 2022. Expert Systems With Applications, Elsevier) 📄