arXiv subjects

Zening Li

Publications and source records attributed to Zening Li.

4 recordsLinked to original sources

arXiv2024

OpenFGL: A Comprehensive Benchmark for Federated Graph Learning

Federated graph learning (FGL) is a promising distributed training paradigm for graph neural networks across multiple local systems without direct data sharing. This approach inherently involves large-scale distributed graph processing, which closely aligns with the challenges and research focuses of graph-based data systems. Despite the proliferation of FGL, the diverse motivations from real-world applications, spanning various research backgrounds and settings, pose a significant challenge to fair evaluation. To fill this gap, we propose OpenFGL, a unified benchmark designed for the primary FGL scenarios: Graph-FL and Subgraph-FL. Specifically, OpenFGL includes 42 graph datasets from 18 application domains, 8 federated data simulation strategies that emphasize different graph properties, and 5 graph-based downstream tasks. Additionally, it offers 18 recently proposed SOTA FGL algorithms through a user-friendly API, enabling a thorough comparison and comprehensive evaluation of their effectiveness, robustness, and efficiency. Our empirical results demonstrate the capabilities of FGL while also highlighting its potential limitations, providing valuable insights for future research in this growing field, particularly in fostering greater interdisciplinary collaboration between FGL and data systems.

cs.LG↗

arXiv2024

Privacy-Preserving Graph Embedding based on Local Differential Privacy

Graph embedding has become a powerful tool for learning latent representations of nodes in a graph. Despite its superior performance in various graph-based machine learning tasks, serious privacy concerns arise when the graph data contains personal or sensitive information. To address this issue, we investigate and develop graph embedding algorithms that satisfy local differential privacy (LDP). We introduce a novel privacy-preserving graph embedding framework, named PrivGE, to protect node data privacy. Specifically, we propose an LDP mechanism to obfuscate node data and utilize personalized PageRank as the proximity measure to learn node representations. Furthermore, we provide a theoretical analysis of the privacy guarantees and utility offered by the PrivGE framework. Extensive experiments on several real-world graph datasets demonstrate that PrivGE achieves an optimal balance between privacy and utility, and significantly outperforms existing methods in node classification and link prediction tasks.

cs.CR↗

arXiv2024

K-stars LDP: A Novel Framework for (p, q)-clique Enumeration under Local Differential Privacy

(p,q)-clique enumeration on a bipartite graph is critical for calculating clustering coefficient and detecting densest subgraph. It is necessary to carry out subgraph enumeration while protecting users' privacy from any potential attacker as the count of subgraph may contain sensitive information. Most recent studies focus on the privacy protection algorithms based on edge LDP (Local Differential Privacy). However, these algorithms suffer a large estimation error due to the great amount of required noise. In this paper, we propose a novel idea of k-stars LDP and a novel k-stars LDP algorithm for (p, q)-clique enumeration with a small estimation error, where a k-stars is a star-shaped graph with k nodes connecting to one node. The effectiveness of edge LDP relies on its capacity to obfuscate the existence of an edge between the user and his one-hop neighbors. This is based on the premise that a user should be aware of the existence of his one-hop neighbors. Similarly, we can apply this premise to k-stars as well, where an edge is a specific genre of 1-stars. Based on this fact, we first propose the k-stars neighboring list to enable our algorithm to obfuscate the existence of k-stars with Warner' s RR. Then, we propose the absolute value correction technique and the k-stars sampling technique to further reduce the estimation error. Finally, with the two-round user-collector interaction mechanism, we propose our k-stars LDP algorithm to count the number of (p, q)-clique while successfully protecting users' privacy. Both the theoretical analysis and experiments have showed the superiority of our algorithm over the algorithms based on edge LDP.

cs.CR↗

arXiv2022

Performance of Cooperative Detection in Joint Communication-Sensing Vehicular Network: A Data Analytic and Stochastic Geometry Approach

The increasing complexity of urban environments introduces additional uncertainty to the deployment of the autonomous vehicular network. A novel road infrastructure cooperative detection model using Joint Communication and Sensing (JCS) technology is proposed in this article to simultaneously achieve high-efficient communication and obstacle detection for urban autonomous vehicles. To suppress the performance fluctuation caused by shadowing and obstruction to the JCS signals, we first derive the statistic of road obstacles from the Geographic Information System (GIS). Then, the analysis of JCS channel characteristics and shadowing factors are presented using Line-of-Sight and Non-Line-of-Sight (LoS and NLoS) channel models under the complex urban scenario. A stochastic geometry approach is applied to analyze the interference factors and the probability distribution of successful JCS detection and communication. Simulations have been made to verify the cooperative detection model by probability analysis based on LoS and NLoS channels, and the numerical results demonstrate several different optimization methods for the deployment of JCS road infrastructures. Finally, we simulated and analyzed a deployment optimization method for JCS road infrastructures that complied with the standard of urban traffic-spot structure placement.

eess.SP↗