- We have 5 papers accepted by AAAI 2023! Congratulations to Zhenwen, Qiannan, Hongyan, Xiuying and Xiaoting! These papers present our recent study of data-efficient learning from graphs, trustworthy machine learning and natural language understanding. More can be found here.
- NSF Center for Computer-Assisted Synthesis led by Prof. Olaf Wiest is funded! In the 5-year project, PI Dr. Zhang’s team will work on designing novel machine learning models for chemical research problems.
- We have another molecular generation paper got accepted by CIKM 2022. It attempts to build an understandable embedding space for molecular generation.
- We have two papers about Recommendation Systems got accepted by CIKM 2022.
- One designs a novel zero-shot learning model for cold-start news recommendation.
- The other investigates the negative sampling issue in the optimization process of recommendation systems, and proposes a fast dynamic negative sampler for the personalized ranking task.
- More of our recent work in this topic can be found at Recommendation Systems.
- Dr. Zhang is appointed as the Editor-in-Chief of ACM SIGKDD Explorations. Welcome submissions!
- Our work of Few-shot Heterogeneous Graph Learning via Cross-domain Knowledge Transfer is accepted by KDD 2022. Anther few-shot learning on graphs survey paper will appear in IJCAI 2022
- Our work of applying PU-learning on graph completion, Positive-Unlabeled Learning with Adversarial Data Augmentation for Knowledge Graph Completion will also appear in IJCAI 2022
- More of our recent work on graph learning can be found at Learning with Graphs.
- Our work of Understanding the Robustness Against Evasion Attack on Categorical Data is accepted by ICLR 2022. More of our work in this topic can be found at Robust and Trustworthy Machine Learning
- Our work of Graph Alignment with Noisy Supervision is accepted by TheWebConf 2022. A related work of handling noisy labels in knowledge graph alignment can be found in our KDD 2020 paper, entitled REA: Robust Cross-lingual Entity Alignment Between Knowledge Graphs.
- Our work of Self-Augmented Graph Contrastive Learning is accepted by AAAI 2022. More of our graph learning work can be found at Learning with Graphs/Networks
- Our work of GNN-Retro: Retrosynthetic planning with Graph Neural Networks is accepted by AAAI 2022. This paper uses GNN to predict the synthesis routes for target molecules. Another related work of molecular graph generation can be found in our CIKM 2021 paper, entitled GF-VAE: A Flow-based Variational Autoencoder for Molecule Generation.
Prof. Zhang joined the department of Computer Science and Engineering at University of Notre Dame in August 2021.