Welcome to UNIST Data Intelligence Lab!Â
We are looking for self-motivated and passionate graduate students and undergraduate interns to join our lab. We are currently interested in the following four areas. If you are interested, please send your transcript, CV, and research interests to Prof. Yeon-Chang Lee (yeonchang at unist.ac.kr).
Trustworthy Graph Neural Networks, which aim to maintain the accuracy of the results while exploring the trustworthiness of the results, particularly in terms of fairness and robustness;
Dynamic Graph Neural Networks, which address the evolving nature of the graph structure and node/edge features over time;
Multi-modal Recommender Systems, which jointly leverage multimodal content of users and items, such as user ratings, product images, and textual descriptions, to achieve high user satisfaction; and
Graph + X, which addresses anomaly detection, career trajectory prediction, etc.
Research Summary
Graph Machine Learning
1) Graph Representation Learning: signed (SIGIR'20, CIKM'20short, SIGIR'23short) and directed (WSDM'22b) graphs
2) Dynamics in Graph Mining: temporal knowledge (ICDM'22, Inf. Sci.'24) and dynamic signed (CIKM'23) graph modeling, and information pathway prediction (KDD'23), and visualization of trajectories on node embeddings
3) Robustness in Graph Mining: out-of-distribution generalization (WWW'23)
4) Fairness in Graph Mining: fair graph neural networks and fair graph anomaly detection
5) Graph-based Downstream Tasks: node ranking (SIGIR'21) and community detection (ICDM'21, TKDE'23)
Recommender Systems
1) Graph-based Collaborative Filtering (AAAI'18, WSDM'22a, SIGWEB Newsletter'23, CSUR'24, IJCAI'24)
2) Multi-modal Recommender Systems (CIKM'22)
3) Domain-specific Recommender Systems: TV shows (ICDE'19, ICDE'23), news articles (ICDE'22industry, CIKM'22short), and related searches
Graph + X: career trajectory prediction, anomaly detection, counter misinformation, etc