Advancements in Privacy-Preserving Collaborative Learning for Graphs

Dear Colleagues,

Graph data provide a powerful paradigm for modeling complex, structured, and interdependent information widely present in domains such as social networks, bioinformatics, transportation systems, recommendation platforms, finance, and smart cities. With the advent of large-scale graph neural networks (GNNs) and graph-based large language models (Graph LLMs), there has been a surge of interest in learning expressive representations that can capture both attribute information and intricate relational dependencies. These advances hold significant potential for enabling intelligent applications across science, industry, and society.

At the same time, the sensitive nature of graph data—ranging from patient records and financial transactions to communication and industrial IoT graphs—raises critical concerns about privacy, security, and compliance. Federated learning has emerged as a promising paradigm to enable collaborative training without centralizing data, thereby mitigating privacy risks. However, new challenges arise when applying federated learning to graph data, such as handling non-IID and heterogeneous distributions, integrating multimodal and relational features, ensuring robustness against adversarial or backdoor attacks, and balancing privacy preservation with model utility. Furthermore, integrating blockchain technologies offers new opportunities for transparent, auditable, and decentralized management of data sharing and model updates while also introducing new research questions on scalability, efficiency, and system design.

This Special Issue aims to bring together cutting-edge research at the intersection of graph data, federated learning, graph LLMs, and blockchain-enabled privacy-preserving training. We welcome high-quality original research articles, comprehensive surveys, and visionary position papers that advance the theoretical foundations, algorithmic innovations, system architectures, and practical applications of privacy-preserving distributed graph learning. Topics of interest include, but are not limited to, the following:

• Privacy-preserving federated learning for graph-structured and multimodal data;
• Techniques for protecting sensitive relationships in graph neural networks;
• Graph unlearning methods;
• Federated training of graph large language models (Graph LLMs);
• Secure aggregation, differential privacy, and homomorphic encryption for graph learning;
• Blockchain-based mechanisms for decentralized trust, auditability, and incentive design in federated graph learning;
• Continual and open-world graph learning with privacy guarantees;
• Applications in domains such as healthcare, finance, social networks, transportation, communication, energy, and industrial IoT;
• Benchmarks, datasets, and evaluation frameworks for privacy-preserving graph learning.

Dr. Zhenzhen Xie
Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Mathematics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• collaborative learning
• graph learning
• privacy