Jun Hong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Jun Hong has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Weiru Liu
University of Bristol
5 shared publications - Jingyu Hu
University of Bristol
3 shared publications - Hongbo Bo
University of Bristol
2 shared publications - Phil Legg
University of the West of England, Bristol, UK
2 shared publications - Sadegh Bamohabbat Chafjiri2 shared publications
- Aisling Gough
Centre for Public Health, School of Medicine, Dentistry and Biomedical Sciences, Queen's University Belfast, Belfast, UK
1 shared publication
Latest Publications
- Open Access
- 27 June 2026
- Published by MDPI AG in Algorithms
- Vol. 19 (7) , 517
Mitigating Degree Bias Adaptively with Hard-to-Learn Nodes in Graph Contrastive Learning
Graph Neural Networks (GNNs) often suffer from degree bias in node classification tasks, where prediction performance varies across nodes with different degrees. Several approaches, which adopt Graph Contrastive Learning (GCL), have been proposed to mitigate this bias. However, the limited number of positive pairs and the equal weighting of all positives and negatives in GCL still lead to low-degree nodes acquiring insufficient and noisy information. This paper proposes the Hardness Adaptive Reweighted (HAR) contrastive loss to mitigate degree bias. It adds more positive pairs by leveraging node labels and adaptively weights positive and negative pairs based on their learning hardness. In addition, we develop an experimental framework named SHARP extending HAR to a broader range of scenarios. Both our theoretical analysis and experiments validate the effectiveness of SHARP. Across four datasets, SHARP outperforms baselines in 14 of 16 settings, improving global accuracy on Cora by 3.6% and 4.2% under the GCN and GAT backbones, while at the degree level it raises accuracy for the lowest-degree nodes by over 10%, confirming that the gains are targeted at the low-degree nodes most affected by degree bias.
- 27 June 2026
- Published by MDPI AG in Algorithms
- Vol. 19 (7) , 517
Open AccessMitigating Degree Bias Adaptively with Hard-to-Learn Nodes in Graph Contrastive Learning
Graph Neural Networks (GNNs) often suffer from degree bias in node classification tasks, where prediction performance varies across nodes with different degrees. Several approaches, which adopt Graph Contrastive Learning (GCL), have been proposed to mitigate this bias. However, the limited number of positive pairs and the equal weighting of all positives and negatives in GCL still lead to low-degree nodes acquiring insufficient and noisy information. This paper proposes the Hardness Adaptive Reweighted (HAR) contrastive loss to mitigate degree bias. It adds more positive pairs by leveraging node labels and adaptively weights positive and negative pairs based on their learning hardness. In addition, we develop an experimental framework named SHARP extending HAR to a broader range of scenarios. Both our theoretical analysis and experiments validate the effectiveness of SHARP. Across four datasets, SHARP outperforms baselines in 14 of 16 settings, improving global accuracy on Cora by 3.6% and 4.2% under the GCN and GAT backbones, while at the degree level it raises accuracy for the lowest-degree nodes by over 10%, confirming that the gains are targeted at the low-degree nodes most affected by degree bias.
- Open Access
- 20 April 2026
- Published by Elsevier in Information and Software Technology
- Vol. 196
Advancing fuzzing with unbiased random generator and Feistel network-based mutations
- 20 April 2026
- Published by Elsevier in Information and Software Technology
- Vol. 196
Open AccessAdvancing fuzzing with unbiased random generator and Feistel network-based mutations
- Open Access
Mitigating Degree Bias Adaptively with Hard-to-Learn Nodes in Graph Contrastive Learning
Graph Neural Networks (GNNs) often suffer from degree bias in node classification tasks, where prediction performance varies across nodes with different degrees. Several approaches, which adopt Graph Contrastive Learning (GCL), have been proposed to mitigate this bias. However, the limited number of positive pairs and the equal weighting of all positives and negatives in GCL still lead to low-degree nodes acquiring insufficient and noisy information. This paper proposes the Hardness Adaptive Reweighted (HAR) contrastive loss to mitigate degree bias. It adds more positive pairs by leveraging node labels and adaptively weights positive and negative pairs based on their learning hardness. In addition, we develop an experimental framework named SHARP to extend HAR to a broader range of scenarios. Both our theoretical analysis and experiments validate the effectiveness of SHARP. The experimental results across four datasets show that SHARP achieves better performance against baselines at both global and degree levels.
Open AccessMitigating Degree Bias Adaptively with Hard-to-Learn Nodes in Graph Contrastive Learning
Graph Neural Networks (GNNs) often suffer from degree bias in node classification tasks, where prediction performance varies across nodes with different degrees. Several approaches, which adopt Graph Contrastive Learning (GCL), have been proposed to mitigate this bias. However, the limited number of positive pairs and the equal weighting of all positives and negatives in GCL still lead to low-degree nodes acquiring insufficient and noisy information. This paper proposes the Hardness Adaptive Reweighted (HAR) contrastive loss to mitigate degree bias. It adds more positive pairs by leveraging node labels and adaptively weights positive and negative pairs based on their learning hardness. In addition, we develop an experimental framework named SHARP to extend HAR to a broader range of scenarios. Both our theoretical analysis and experiments validate the effectiveness of SHARP. The experimental results across four datasets show that SHARP achieves better performance against baselines at both global and degree levels.
- Open Access
ProxiMix: Enhancing Fairness with Proximity Samples in Subgroups
Many bias mitigation methods have been developed for addressing fairness issues in machine learning. We found that using linear mixup alone, a data augmentation technique, for bias mitigation, can still retain biases present in dataset labels. Research presented in this paper aims to address this issue by proposing a novel pre-processing strategy in which both an existing mixup method and our new bias mitigation algorithm can be utilized to improve the generation of labels of augmented samples, which are proximity aware. Specifically, we proposed ProxiMix which keeps both pairwise and proximity relationships for fairer data augmentation. We conducted thorough experiments with three datasets, three ML models, and different hyperparameters settings. Our experimental results showed the effectiveness of ProxiMix from both fairness of predictions and fairness of recourse perspectives.
Open AccessProxiMix: Enhancing Fairness with Proximity Samples in Subgroups
Many bias mitigation methods have been developed for addressing fairness issues in machine learning. We found that using linear mixup alone, a data augmentation technique, for bias mitigation, can still retain biases present in dataset labels. Research presented in this paper aims to address this issue by proposing a novel pre-processing strategy in which both an existing mixup method and our new bias mitigation algorithm can be utilized to improve the generation of labels of augmented samples, which are proximity aware. Specifically, we proposed ProxiMix which keeps both pairwise and proximity relationships for fairer data augmentation. We conducted thorough experiments with three datasets, three ML models, and different hyperparameters settings. Our experimental results showed the effectiveness of ProxiMix from both fairness of predictions and fairness of recourse perspectives.
- 22 August 2024
- Published by Springer Nature
- p. 143-160
AdaHAT: Adaptive Hard Attention to the Task in Task-Incremental Learning
- 22 August 2024
- Published by Springer Nature
- p. 143-160
AdaHAT: Adaptive Hard Attention to the Task in Task-Incremental Learning
- Open Access
- 19 May 2024
- Published by Elsevier in Computers & Security
- Vol. 143
Vulnerability detection through machine learning-based fuzzing: A systematic review
- 19 May 2024
- Published by Elsevier in Computers & Security
- Vol. 143
Open AccessVulnerability detection through machine learning-based fuzzing: A systematic review
- Top Cited PapersOpen Access
- 23 March 2017
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 3 (1) , e14
Tweet for Behavior Change: Using Social Media for the Dissemination of Public Health Messages
A multidisciplinary journal that focuses on public health and technology, public health informatics, mass media campaigns, surveillance, and innovation in public health practice and research. Also dedicated to rapid open data sharing during epidemics.
- 23 March 2017
- Published by JMIR Publications Inc. in JMIR Public Health and Surveillance
- Vol. 3 (1) , e14
Top Cited PapersOpen AccessTweet for Behavior Change: Using Social Media for the Dissemination of Public Health Messages
A multidisciplinary journal that focuses on public health and technology, public health informatics, mass media campaigns, surveillance, and innovation in public health practice and research. Also dedicated to rapid open data sharing during epidemics.
- Open Access
- 30 August 2016
- Published by Springer Nature
- p. 214-229
A Two-Stage Online Approach for Collaborative Multi-agent Planning Under Uncertainty
- 30 August 2016
- Published by Springer Nature
- p. 214-229
Open AccessA Two-Stage Online Approach for Collaborative Multi-agent Planning Under Uncertainty