The New Special Issue "Entropy-Centric Intelligent Computation with Graph: In Pursuit of Advanced Computational Theories, Methods, and Applications

Guest Editors: Dr. Yongpan Sheng, Dr. Hao Wang, Dr. Junyang Chen, and Dr. Chunwei Tian

Submit to the Special Issue: https://www.mdpi.com/si/entropy/6837W6XEIV

Submission deadline: 20 November 2025

Special Issue Information:

Entropy-centric intelligent computation is essential to dealing with diverse forms of graph computation issues, such as link prediction, graph classification, graph matching, graph structure learning, graph generation, and transformation. This method can be applied in a variety of different settings, including knowledge graphs, social networks, spatio-temporal networks, IoT sensing, recommendation system, self-driving cars, bioinformatics, and medical informatics.

Recently, structural entropy-based methods have been adopted to efficiently rank graph nodes. Likewise, graph entropy-based methods are employed to automatically embed dimension selection when learning different types of graph representation from the perspective of the minimum entropy principle. Furthermore, there has been a recent ware of information bottleneck-based methods developed to optimally balance the expressiveness and robustness of the learned graph representation, recognize predictive graph substructures, conduct highly efficient graph training with optimizing adversarial graph augmentation strategies, etc.

However, despite these successes, as a promising entropy-centric graph analysis and computing paradigm, significantly challenging issues still exist, such as how to model existing diverse graph structure-based data that often represent multi-modal, multi-relational, and dynamic graphs from the perspective of the entropy principle, or how to efficiently learn the neural graph representation of the large-scale field-specific graph to use its rich structural and semantic information to guide the entropy principle. Besides, it is also necessary to investigate how new entropy-centric computing theories, technologies, and novel applications might be integrated into the current and future graph computation framework.

This Special Issue will be a forum for researchers working on mining and learning from entropy-centric intelligent computation with graphs in pursuit of advanced computational theories, methods, and applications. Submitted research papers and comprehensive reviews should focused on the following research areas:

• Entropy-centric intelligent computation theories with graphs;
• Entropy-centric graph structured-based data modeling with time-evolving, multi-relational, and multi-modal nature;
• Neural graph representation learning for homogeneous or heterogeneous graphs in the guidance of the entropy principle;
• Entropy-centric data mining for knowledge graphs, linguistics graphs, bibliographic graphs, textual graphs, social networks, traffic networks, and molecules;
• New entropy-centric computing framework/method for graph structure-based data;
• Applications of entropy-centric graph mining in e-commerce, text mining, stock prediction, recommendation systems, self-driving cars, protein modeling, program analysis, etc.