Intelligent Software Laboratory

Notice!

We are recruiting!
  • MS course (석사과정)
  • Ph.D course (박사과정)
  • MS + Ph.D integrated course (통합과정)
  • Undergraduated Research Internship (partime/fulltime) (학부연구생)
If you want to study AI and make something fun and useful for our life, apply to here or e-mail me.

We seek to utilize artificial intelligence(AI) to design, research, and implement the future of our lives. In particular, we study how AI can be used to reduce unnecessary and repetitive tasks in order to maximize human creative ability.

To accomplish the goals, We are studying the following major topics:

AI for everyday-life

  • Natural Language Understanding
  • Dialog Management System including Chatbot
  • Human-Machine, Machine-Machine dialog system

AI for creator

  • Multimedia(music, image, video) generation
  • Interactive content refinment

AI for student

  • Machine Comprehension and Answering
  • Multimodal understanding

User Understanding

  • Natural Language Processing
  • Text query and sequential behavior understanding
  • User Modeling
  • Simulated User

Work Experience

  • SK telecom, 2014~2018
  • ETRI, 2012~2014
  • Samsung Electronics, 2010~2012

Projects

AI for creator

Multimedia creation AI

AI for everyday-life

Natural Language Understanding using Deep Learning, Everyday-life powerup-tools

AI for student

Design and implement the future of study and education with AI

User Understanding

User understanding using Deep Learning

Selected Publications

Anaphora resolution with pointer networks

PDF

Hybrid User Intention Modeling to Diversify Dialog Simulations

PDF

Example-based dialog modeling for practical multi-domain dialog system

PDF

Recent Publications

More Publications

underline ( ) : corresponding author

. Semantic Vector Learning for Natural Language Understanding. In Computer Speech & Language, Accepted (To be appeared), 2018.

. Multitask Pointer Network for Korean Dependency Parsing. In Transactions on Asian and Low-Resource Language Information Processing, Accepted (To be appeared), 2018.

. Effective Korean units for sequence encoding in deep learning. In Journal of KIISE, Vol 45 no. 05, Pages 0457 ~ 0465, May, 2018.

PDF

. Deep Neural Architecture for Recovering Dropped Pronouns in Korean. In ETRI Journal, Volume 40, Pages 257-265, 11 April, 2018.

PDF

. End-to-End Korean Part-of-Speech Tagging Using Copying Mechanism. TALLIP, 2018.

PDF

. Concept Equalization to Guide Correct Training of Neural Machine Translation. In IJCNLP, 2017.

PDF

. Anaphora resolution with pointer networks. In Pattern Recognition Letters, Volume 95, Pages 1-7, 1 August, 2017.

PDF

. Refining sentence similarity with discourse information in dialog system. In INTERSPEECH, 2013.

PDF

. Hybrid User Intention Modeling to Diversify Dialog Simulations. In Computer Speech and Language, Volume 25 Issue 2, Pages 307-326, April, 2011.

PDF

. Hybrid approach to robust dialog management using agenda and dialog examples. In Computer Speech and Language, Volume 24 Issue 4, Pages 609-631, October, 2010.

PDF

Selected Patents

  • 정상근, 이청재, 이근배. 음성대화 오류검증을 통한 확인대화 방법 및 장치. 대한민국 특허등록 (등록번호 제 10-0732611호, 2007-06-20)
  • 이청재, 정상근, 이근배. 대화관리 장치 및 그를 위한 대화예제 기반의 대화 모델링 기법을 통한 대화관리 방법. 대한민국 특허등록(등록번호 제 10-0772660호, 2007-10-26)
  • [프로그램 등록] 김경덕, 이근배, 이동현, 이청재, 정상근, 스마트홈 제어를 위한 멀티모달 대화 인터페이스. 대한민국 프로그램 등록 (등록번호 2009-01-221-005681, 2009-10-23)
  • 이성진, 정상근, 김경덕, 이청재, 이근배. 외국어 회화 연습 방법 및 장치와 이것의 빠른 개발과 검증을 돕는 학습자 시뮬레이션 방법과 장치. 대한민국 특허등록(등록번호 제 10-1037247호, 2011-05-19)
  • 김경덕, 정상근, 이청재, 이근배. 데이터 수집 시스템 및 방법. 대한민국 특허 등록(등록번호 제 10-1093311호, 2011-12-06)

Teaching

I am a teaching instructor for the following courses at Chungnam National University:

2018 Fall

  • Linear Algebra
  • Obejct Oriented Design

Contact

  • hugman@cnu.ac.kr
  • 82 042 821 5444
  • W2-624, Daehak-ro, Yuseong-gu, Daejeon 34141, Chungnam National University, Republic of Korea
  • Wednesday 13:00 to 15:00 or email for appointment

Recent Posts