O-Joun Lee

Catholic University of Korea, Artificial Intelligence, Faculty Member

Followers

Following

Public Views

O-Joun Lee has been an assistant professor at the Catholic University of Korea, Rep. of Korea since September 2021. He received Ph.D. degree in Computer Science and Engineering from Chung-Ang University in 2019. He also received B.Eng. degree in Software Science from Dankook University in 2015. He was a full-time researcher at Pohang University of Science and Technology, Rep. of Korea from September 2019 to August 2021. His research interests include graph mining, graph representation learning, and graph neural networks. He has applied graph mining models and methods to various unstructured data, such as social media, bibliographic data, medical knowledge bases, traffic flow data, molecular structures, etc.

less

Related Authors

Interests

Uploads

Papers by O-Joun Lee

Pre-training Graph Neural Networks on Molecules by Using Subgraph-Conditioned Graph Information Bottleneck

Proceedings of the 39th AAAI Conference on Artificial Intelligence (AAAI 2025), 2025

This study aims to build a pre-trained Graph Neural Network (GNN) model on molecules without huma... more This study aims to build a pre-trained Graph Neural Network (GNN) model on molecules without human annotations or prior knowledge. Although various attempts have been proposed to overcome limitations in acquiring labeled molecules, the previous pre-training methods still rely on semantic subgraphs, i.e., functional groups. Only focusing on the functional groups could overlook the graph-level distinctions. The key challenge to build a pre-trained GNN on molecules is how to (1) generate well-distinguished graph-level representations and (2) automatically discover the functional groups without prior knowledge. To solve it, we propose a novel Subgraph-conditioned Graph Information Bottleneck, named S-CGIB, for pre-training GNNs to recognize core subgraphs (graph cores) and significant subgraphs. The main idea is that the graph cores contain compressed and sufficient information that could generate well-distinguished graph-level representations and reconstruct the input graph conditioned on significant subgraphs across molecules under the S-CGIB principle. To discover significant subgraphs without prior knowledge about functional groups, we propose generating a set of functional group candidates, i.e., ego networks, and using an attention-based interaction between the graph core and the candidates. Despite being identified from self-supervised learning, our learned subgraphs match the real-world functional groups. Extensive experiments on molecule datasets across various domains demonstrate the superiority of S-CGIB.

DownloadEdit

Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-Based Similarity

Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI 2024), 2024

Graph representation learning (GRL) methods, such as graph neural networks and graph transformer ... more Graph representation learning (GRL) methods, such as graph neural networks and graph transformer models, have been successfully used to analyze graph-structured data, mainly focusing on node classification and link prediction tasks. However, the existing studies mostly only consider local connectivity while ignoring long-range connectivity and the roles of nodes. In this paper, we propose Unified Graph Transformer Networks (UGT) that effectively integrate local and global structural information into fixed-length vector representations. First, UGT learns local structure by identifying the local sub-structures and aggregating features of the k-hop neighborhoods of each node. Second, we construct virtual edges, bridging distant nodes with structural similarity to capture the long-range dependencies. Third, UGT learns unified representations through self-attention, encoding structural distance and p-step transition probability between node pairs. Furthermore, we propose a self-supervised learning task that effectively learns transition probability to fuse local and global structural features, which could then be transferred to other downstream tasks. Experimental results on real-world benchmark datasets over various downstream tasks showed that UGT significantly outperformed baselines that consist of state-of-the-art models. In addition, UGT reaches the third-order Weisfeiler-Lehman power to distinguish non-isomorphic graph pairs.

DownloadEdit

O-Joun Lee

Uploads

Papers by O-Joun Lee

Log In