Minsang Kim


I am machine learning research engineer at SK Telecom.

Research Interest: NLP, LLM, Data-Centric AI, and Foundation Models

Mail: kmswin7 [at] gmail [dot] com

CV  /  LinkedIn  /  Google Scholar  /  Github

profile photo
Publications
HPLC Hierarchical Position Embedding of Graphs with Landmarks and Clustering for Link Prediction
Minsang Kim, Seungjun Baek
WWW, 2024  
arXiv / github

ComDensE ComDensE : Combined Dense Embedding of Relation-aware and Common Features for Knowledge Graph Completion
Minsang Kim, Seungjun Baek
ICPR, 2022  
arXiv / github

PKG Persona-Knowledge Dialogue Multi-Context Retrieval and Enhanced Decoding Methods
Minsik Oh* Minsang Kim* (* : equal contributions)
COLING-PKG, 2022  
arXiv / github

OASYS OASYS: Domain-Agnostic Automated System for Constructing Knowledge Base from Unstructured Text
Minsang Kim*, Sanghyun-Je*, Eunjoo Park (* : equal contributions)
SIGKDD-MLG, 2022  
arXiv / github

Awards
pkg-workshop 1st place award in both retrieval and generation shared tasks
COLING 2022 workshop on Customized Chat Grounding Persona and Knowledge
Work Experience
SKT Machine Learning Research Engineer | SK Telecom.
Seoul, South Korea | January, 2023 ~ Current
  • Modeling Large-scale language models.
  • Developing Korean ChatGPT.
  • Modeling Finetuning pipeline of Large-scale language models.
KEP Machine Learning Research Engineer | Kakao Enterprise.
Pangyo, Gyeonggi-do, South Korea | December, 2019 ~ December, 2022
  • Modeling NLU (Assistant, Bot classifier, Intent Classifier, Machine Learning Entity Tagger), Open-domain QA tasks for Kakao-i assistant service
  • Modeling Network analysis for B2B service
  • Modeling automatic knowledge graph construction system for B2B service
  • Modeling personalized member search ranking for B2B messenger service
FastCampus Lecturer | Fast campus.
Gangnam-gu, Seoul, South Korea | September, 2022 ~ November 2022
  • I lectured on several NLP categories - Machine translation, Sentence classification, Passage retrieval, Dialogue system, Knowledge graph, and Information extraction. Lecture link
Kakao Machine Learning Research Engineer | Kakao Corp.
Pangyo, Gyeonggi-do, South Korea | July, 2019 ~ December, 2019
  • Develop Daum Search Engines
Extra-Curricular Activities
draper KIC college R2BD (Research to Business Development) Program | Draper University, San Mateo & Industrial Technology research institute, Berlin
San Mateo, California, U.S.A & Berlin, Germany | July, 2017 ~ August, 2017

In the summer of 2017, I attended to KIC college R2BD program, which was participated by several universities in South Korea, and I was on behalf of our school for cultivating entrepreneurship. I and My colleague introduced a pill scanner that classifies what kind of pills by using a deep learning model and then made a business plan for this project. During this program, I learned about teamwork, design-thinking, presentation skills, etc. from various activities. Furthermore, I had progressed our business plan combined with methods that were taught from various activities. Finally, I presented my project result in front of a lot of international students and Tim Draper who is a famous venture capital investor in Silicon Valley.

This website is adapted from Jon Barron's template.