Special Session on
Deep Graph Clustering Models

Aim and Scope

The primary goal of this session is to bring together researchers and practitioners to explore the latest advancements in Deep Graph Clustering (DGC). While traditional clustering focuses on data in Euclidean space, DGC leverages the power of Graph Neural Networks (GNNs) to handle complex relational data.

This session aims to address the fundamental challenges of DGC, such as:

• Integrating node attributes with structural topology.
• Developing self-supervised and unsupervised objectives that avoid collapsed solutions.
• Scaling deep clustering algorithms to massive, real-world graphs.

Deep Graph Clustering represents a fusion of unsupervised learning and graph representation learning. Unlike standard GNNs that require labels, DGC models learn a low-dimensional embedding space where nodes belonging to the same community are positioned closely together, typically optimized via a joint clustering and embedding loss.

Main Topics of Interest Include:

• Autoencoder-based Models: Graph Autoencoders (GAE) and Variational Graph Autoencoders (VGAE) for clustering.
• Contrastive Learning for Graphs: Recent shifts toward self-supervised contrastive objectives (e.g., node-to-node or node-to-subgraph) to improve cluster discriminability.
• Hybrid Architectures: Combining Graph Convolutional Networks (GCNs) with traditional clustering algorithms like K-Means or Gaussian Mixture Models (GMMs) in an end-to-end framework.
• Application-Specific DGC: Deploying these models in bioinformatics (protein-protein interaction), social network analysis, and recommendation systems.
• Dynamic and Heterogeneous Graphs: Clustering techniques for graphs that evolve over time or contain multiple types of nodes and edges.

Paper Submission

Potential authors may submit their manuscripts for presentation consideration through ICETCI 2026 submission system electronically at https://edas.info/N34670 , following the conference guidelines. All submissions will go through peer review process. To submit your paper to this special session, you have to choose our special session title on the submission page.

Important Dates

Organizers

Dr. Panthadeep Bhattacharjee received his PhD. in Computer Science and Engineering from the Indian Institute of Technology Guwahati (IITG). His broad area of research includes Data Mining, Machine Learning, Pattern Recognition, Incremental Algorithms. He had obtained his M.Tech in Computer Science and Engineering from the National Institute of Technology, Durgapur (NITD). In course of his masters he had developed an optimal virtual keyboard in Bengali language for faster text entry rate. Dr. Panthadeep had completed his B.Tech in Information Technology from Assam (Central) University, Silchar. He had studied his higher secondary from the prestigious Cotton University (formerly Cotton College) Guwahati. Prior to joining NIT Rourkela, Dr. Bhattacharjee was working as an Assistant Professor in the Department of Computer Science and Engineering, Indian Instituite of Information Technology Guwahati (IIITG). He had also served as an Assistant Professor in the School of Computer Engineering, KIIT Deemed to be University, Bhubaneswar. Dr. Panthadeep has been a professional member of the Association of Computing Machinery (ACM) since 2016. He had received the best thesis award in the PhD. research Symposium track as a part of the ICDCIT 2022 conference. He had primarily worked in developing incremental algorithms for handling dynamic data for the purpose of mining clusters and outliers. As a part of his research, Dr. Bhattacharjee had conducted one of the most comprehensive studies about density-based clustering algorithms that were proposed since 1996 onwards ranging over a period of 23 years. He was also offered a position of Post-Doctoral researcher in the University of Porto, Portugal for working in a project titled "Generating Artificial Data for ML algorithms".

Email: bhattacharjeep@nitrkl.ac.in

Dr. Priodyuti Pradhan is an Assistant Professor in the Department of Computer Science and Engineering at IIIT Raichur, which he joined in November 2022. He earned his Ph.D. in Network Science from the Indian Institute of Technology (IIT) Indore. Before joining IIIT Raichur, he served as an Assistant Professor at the University of Petroleum and Energy Studies (UPES), Dehradun, and was a Postdoctoral Research Fellow at the Complex Networks Dynamics Lab, Bar-Ilan University, Israel.
His research expertise is centered on Network Science, specifically focusing on information propagation patterns in complex networks, nonlinear dynamics, and Graph Neural Networks (GNNs). He is also actively exploring machine learning-based data-driven modeling of complex systems and time series data analysis. Dr. Pradhan has published significant work in journals such as Physical Review E and Chaos, Solitons & Fractals, and leads the networks.ai Lab at IIIT Raichur.

Email: prio@iiitr.ac.in

Dr. Sandeep Vidyapu is a researcher at the Visualization Research Center (VISUS) of the University of Stuttgart, Germany. He earned his Ph.D. from the Indian Institute of Technology (IIT) Guwahati. Dr. Vidyapu’s work is particularly relevant to deep learning as he specializes in "Visual XAI," developing methods to interpret and curate complex data structures like scene graphs. He has a strong track record of publications in premier venues such as ACM Transactions on the Web (TWEB) and the Symposium on Eye Tracking Research and Applications (ETRA). His current research involves using visualization techniques to explain the decision-making processes of graph-based models, such as Visual Question Answering (VQA) systems.

Email: sandy.apj911@gmail.com