I’m PhD student at Pohang university of Science & Technology (POSTECH).

ryusangwon@postech.ac.kr

Research Interest

• Natural Language Processing
• Large Language Model
• Text Summarization
• Reinforcement Learning
• Data Mining

Education

• POSTECH
  • Ph.D. student in Graduate School of Artifical Intelligence
    Advised by Prof. Gary Geunbae Lee and Jungseul Ok
    Thanks to Prof. Yunsu Kim
• Ajou University
  • B.S. in Software and Computer Engineering
  • Micro Major in Artifical Intelligence

Research Experiences

• Research Assistant, POSTECH, NLP Group, February 2023 - current
  Advised by Prof. Yunsu Kim and Prof. Gary Geunbae Lee
• Research Intern, Ajou University, Lamda Lab, February 2022 - December 2022
  Advised by Prof. Kyung-Ah Sohn
• Research Intern, Ajou University, HMS Lab, February 2021 - December 2021
  Advised by Prof. Sangeun Oh

Publications

*: Equal contributions

conferences

• Autoregressive Multi-trait Essay Scoring via Reinforcement Learning with Scoring-aware Multiple Rewards
  Heejin Do, Sangwon Ryu, Gary Geunbae Lee
  EMNLP 2024 (long), November 2024 [paper]
• Key-Element-Informed sLLM Tuning for Document Summarization
  Sangwon Ryu*, Heejin Do*, Yunsu Kim, Gary Geunbae Lee, Jungseul Ok
  INTERSPEECH 2024, September 2024 [paper] (Oral presentation)
• Multi-Dimensional Optimization for Text Summarization via Reinforcement Learning
  Sangwon Ryu*, Heejin Do*, Yunsu Kim, Gary Geunbae Lee, Jungseul Ok
  ACL 2024 (long), August 2024 [paper]
• Comparative Analysis of Language Model Performance in News Domain Summarization
  Sangwon Ryu, Yunsu Kim, Gary Geunbae Lee
  HCLT 2023, October 2023 (Oral presentation)
• Query Control for Fluent Query Focused Summarization
  Sangwon Ryu, Kyung-Ah Sohn
  KCSE 2023, February 2023 (Best paper award, Oral presentation)
• NLI Style few-shot learning with supervised contrastive loss
  Sangwon Ryu, Kyung-Ah Sohn
  KICS 2022, June 2022

preprints

• Ontology-Free General-Domain Knowledge Graph-to-Text Generation Dataset Synthesis using Large Language Model
  DaeHee Kim, Deokhyung Kang, Sangwon Ryu, Gary Geunbae Lee
  arxiv, September 2024 [paper]

Teaching Experiences

• Mentor, POSTECH, Undergraduate Research Participation
  • Text Summarization, February 2024 - June 2024
• Lecturer, POSCO AI Application Expert Training
  • Data Mining, April 2024 - August 2024
  • Data Mining, April 2023 - August 2023
• Teaching Assistant, Megastudy
  • Mathematics, May 2019 - December 2019

Responsibilities

• Reviewer
  • Conference: ACL (SRW 2024)
  • Journal: TASL (2024)

Research Projects

• MILAINA: Multi-Lingual AI News Agent
  National Research Foundation of Korea (NRF), Principal Investigator, June 2023 - current
• Development and Verification of Online Hate Speech Identification Model
  National Research Foundation of Korea (NRF), July 2022 - December 2022

Application Projects

• HighwayLoL (LOL mento-menti Service) (Backend), August 2022 - December 2022
  • Technology used: Javascript, Nodejs, SQL, Docker, Oracle Cloud
• Mobile Cross Platform UI Distribution (Android), April 2021 - December 2021
  • Technology used: Java, Android, Soot

Awards and Certificates

• Best Paper Award, KCSE (2023)
• SOFTCON Research Excellence Award, Ajou University (2022)
• Big Data Analysis Practice Level 2
• Recognition Award, Gyeonggi Nambu Provincial Police Agency (2017)
• Excellence Recruit, Korea Army (2016)

Patents and Copyrights

• Instruction Tuned sLLM for News Summarization
  Copyright, Registration Number: C-2024-000076
• sLLM Tuning method for News Summarization, device for the same, and method for generating news summarization
  Patent, Registration Number: 10-2023-0191656
• Query focus summarization Program
  Copyright, Registration Number: C-2022-052465

Skills

• Python, Pytorch, (vim, Nodejs, django, SQL, Android)

Other Experiences

• Forklift Driver, September 2018 - December 2018
• Bus Driver, Auxiliary Police, October 2016 - July 2018
• GJJ, White Belt (I)
• Interests in {Body Building, MMA, Grappling}