Machine Learning on Graphs (MLoG) Workshop

About

Graphs, which encode pairwise relations between entities, are a kind of universal data structure for a lot of real-world data, including social networks, transportation networks, and chemical molecules. Many important applications on these data can be treated as computational tasks on graphs. Recently, machine learning techniques are widely developed and utilized to effectively tame graphs for discovering actionable patterns and harnessing them for advancing various graph-related computational tasks. Huge success has been achieved and numerous real-world applications have benefited from it. However, since in today’s world we are generating and gathering data in a much faster and diverse way, real-world graphs are becoming increasingly large-scale and complex. More dedicated efforts are needed to propose more advanced machine learning techniques and properly deploy them for real-world applications in a scalable way.

MLoG at ICDM'23 provides a venue to gather the academia researchers and industry researchers/practitioners to present the recent progress of machine learning on graphs.

Topics of Interest

Graphs are a kind of universal data structure for representing pair-wise relationships between entities, which can be ubiquitously observed in different domains ranging from computer science, social science, physics, chemistry, to biology. Many real-world applications can be treated as computational tasks on graphs. For example, friend recommendation in social networks can be regarded as a link prediction task and predicting properties of chemical compounds can be treated as a graph classification task. To facilitate these tasks, machine learning techniques have been widely adopted to perform analysis. As our ability of generating and collecting data constantly increasing in a unprecedented way, the graph-structure data we are facing in the modern era (especially coming from the web) are becoming more and more diverse, complex and large-scale. Hence, more efforts are required for developing more effective algorithms and deploying them efficiently for real-world applications. In this workshop, we aim to discuss the recent research progress of machine learning on graphs in both theoretical foundations and practical applications. We invite submissions that focus on recent advances in research/development of machine learning on graphs along with their applications.

Theory and methodology papers are welcome from any of the following areas, including but not limited to:

Graph Kernels
Graph Summarization
Graph Coarsening
Graph Alignment
Graph Generative Models
Graph Mining
Graph Neural Networks
Network Embedding
Machine Learning for Graph Combinatorial Optimization
Graph Feature Engineering and Selection
Scalable Graph Learning Models and Methods

and application papers focused on but not limited to:

Recommender Systems
Computer Vision
Natural Language Processing
Bioinformatics (e.g., drug discovery)
Cybersecurity (e.g., malware detection/propagation)
Financial security (e.g., fraudster detection)
Transportation/mobility networks (e.g., traffic prediction)
Graph ML Platforms and Systems (e.g., in-database machine learning)

Submission Details

All submissions should be 2 to 8 pages (including all references, tables, and figures), double-column pdfs, and following the IEEE conference format - please refer to the ICDM'23 website for futher details here.

Submissions will be reviewed double-blind, and author names and affiliations should NOT be listed. Submitted works will be assessed based on their novelty, technical quality, potential impact, and clarity of writing (and should be in English). For papers that primarily rely on empirical evaluations, the experimental settings and results should be clearly presented and repeatable. We encourage authors to make data and code available publicly when possible. The best paper (according to the reviewers' ratings) will be announced at the end of the workshop. Following ICDM tradition, all accepted works at the MLoG workshop will be published in formal proceedings by the IEEE Computer Society Press. Therefore, papers must not have been accepted for publication elsewhere or be under review for other workshops (that are archival), conferences, or journals.

To submit your work, please use the following: TBD

Note that at least one of the authors of the accepted workshop papers must register for the workshop (details to come on the main ICDM'23 website). For questions about submission, please contact us at: yao.ma@njit.edu

Workshop Program

Accepted Papers

Dynamic Relation-Attentive Graph Neural Networks for Fraud Detection.
Heehyeon Kim, Jinhyeok Choi, and Joyce Whang

E2EG: End-to-End Node Classification Using Graph Topology and Text-based Node Attributes.
Tu Anh Dinh, Jeroen den Boef, Joran Cornelisse, and Paul Groth

Self-supervision meets kernel graph neural models: From architecture to augmentations.
Jiawang Dan, Ruofan Wu, Yunpeng Liu, Baokun Wang, Changhua Meng, Tengfei Liu, Tianyi Zhang, Ningtao Wang, Xing Fu, Qi Li, and Weiqiang Wang

On Graph Representation based Re-Identification -- A Proof of Concept.
Simon Klüttermann, Jérôme Rutinowski, Anh Nguyen, Christopher Reining, Moritz Roidl, and Emmanuel Müller

Hyperbolic Node Structural Role Embedding.
Lili Wang, Chenghan Huang, Weicheng Ma, Zhongyang Li, and Soroush Vosoughi

Study of Topology Bias in GNN-based Knowledge Graphs Algorithms.
Anil Surisetty, Aakarsh Malhotra, Sudipta Modak, Siddharth Yerramsetty, Alok Singh, Liyana Sahir, Esam Abdel-Raheem, and Deepak Chaurasiya

Clustering with Entropy-based Recombination for Training GCNs on Large Graphs.
Shangwei Wu, Yingtong Xiong, Hui Liang, and Chuliang Weng

Influence Propagation for Linear Threshold Model with Graph Neural Networks.
Francisco Santos, Anna Stephens, Pang-Ning Tan, and Abdol-Hossein Esfahanian

DropMix: Better Graph Contrastive Learning with Harder Negative Samples.
Yueqi Ma, Minjie Chen, and Xiang Li

Aligning Contrastive Clusters for Cross-Network Node Classification.
Linyao Yang, Hongyang Chen, and Xiao Wang

SKGHOI: Spatial-Semantic Knowledge Graph for Human-Object Interaction Detection.
Lijing Zhu, Qizhen Lan, Alvaro Velasquez, Houbing Song, Acharya Kamal, Qing Tian, and Shuteng Niu

Effectiveness of Term Frequency-Inverse Graph Frequency (TF-IGF) Technique Against Various Cyber Attacks.
Prabin Lamichhane, Jacob Taylor, and William Eberle

Exhaustive Evaluation of Dynamic Link Prediction.
Farimah Poursafaei and Reihaneh Rabbany

Exploiting Local Information with Subgraph Embedding for Graph Neural Networks.
Hyung-Jun Moon and Sung-Bae Cho

SubAnom: Efficient Subgraph Anomaly Detection Framework over Dynamic Graphs.
Chi Zhang, Wenkai Xiang, Xingzhi Guo, Baojian Zhou, and Deqing Yang

Continuous-Time Temporal Graph Learning on Provenance Graphs.
Jakub Reha, Giulio Lovisotto, Michele Russo, Alessio Gravina, and Claas Grohnfeldt

Workshop Program. (Local time GMT+8 in China at ICDM'23).

08:00 - 08:05 Welcome and Opening Remarks
08:05 - 08:50 Keynote - Jiezhong Qiu of Zhejiang University
08:50 - 09:35 Keynote - Kijung Shin of Korea Advanced Institute of Science & Technology
09:35 - 10:20 Keynote - Qiaoyu Tan of New York University Shanghai

10:20 - 10:30 Break

10:30 - 11:15 Keynote - Siheng Chen of Shanghai Jiao Tong University
11:15 - 12:00 Keynote - Wenqi Fan of Hong Kong Polytechnic University

12:30 - 12:45 SubAnom: Efficient Subgraph Anomaly Detection Framework over Dynamic Graphs
12:45 - 13:00 DropMix: Better Graph Contrastive Learning with Harder Negative Samples
13:00 - 13:15 Study of Topology Bias in GNN-based Knowledge Graphs Algorithms

13:15 - 13:30 Break

13:30 - 14:00 Poster Session