We conduct various data mining studies focusing on recommender systems with deep learning

• Recommender systems for personalized and group recommendation with various deep learning techniques.

• Data mining for analyzing data from IoT devices to find the hidden knowledge in smart home area.

• Natural language processing for Korean language education. 

한국연구재단 우수신진연구

• 게임이론과 그래프 신경망 기반의 저비용/고효율 그룹 추천 시스템 (2022.03 ~ 2025.02)

• 연구책임자

사람중심 인공지능 핵심 원천기술 개발사업

• 과기정통부 인공지능융합혁신인재양성사업 (2022.07 ~ 2026.12)

인공지능대학원(후속사업) 프로그램

• 과기정통부 인공지능융합혁신인재양성사업 (2022.07 ~ 2025.12)

한국연구재단 학제간융합연구지원사업

• 사용자 중심의 한국어 텍스트 분석 도구 (U-KTA) 개발 (2023.06 ~ 2026.05)

4단계 BK21사업 미래인재 양성사업

• 산업융합형 차세대 인공지능 혁신인재 교육연구단 (2020.09 ~ 2027.08)

Group Recommendation based on Game Theory and GNN

• Improving the efficiency of a group recommender system using low-dimensional user and item features.

• Maximizing the satisfaction for group users through recommendation results based on game theory.

• Group recommender system based on graph neural networks for large groups.

Continual Learning, Knowledge Graph, and GNN for Personalized Recommendation

• Integration of heterogeneous preference information on the personalized recommender system.

• Representing a recommender system and inferring recommendations from a knowledge graph.

• Continual learning for personalized recommender system based on graph neural networks.

Improving the Quality of Sensor Data in Smart Homes

• An efficient calibration framework of low-cost sensors for daily life (SenDaL).

• Detecting implicit information that cannot be physically sensed by a sensor.

• Finding the new knowledge based on monitoring the sensors.

Natural Language Processing for Korean Language Education 

• Development of Transformer-based Korean morpheme analyzer.

• Development of pre-training and fine-tuning models for Korean language education domain.