QoSGMAA: A Robust Multi-Order Graph Attention and Adversarial Framework for Sparse QoS Prediction

TL;DR

QoS prediction in highly sparse, noisy service environments is challenging. The authors propose QoSMGAA, a framework that integrates a multi-order graph attention mechanism with an adversarial interaction module and Gumbel-Softmax sampling to learn robust, high-order user–service relationships. Across real-world datasets, QoSMGAA achieves superior accuracy (lower MAE and RMSE) than CF, MF, DL, and GNN baselines, with pronounced gains in sparse regimes and resilience to structural noise, while maintaining practical inference time and memory footprints. The work demonstrates strong potential for robust service selection and recommendation in large-scale, sparse QoS landscapes and outlines directions such as diffusion-based denoising to further improve robustness.

Abstract

With the rapid advancement of internet technologies, network services have become critical for delivering diverse and reliable applications to users. However, the exponential growth in the number of available services has resulted in many similar offerings, posing significant challenges in selecting optimal services. Predicting Quality of Service (QoS) accurately thus becomes a fundamental prerequisite for ensuring reliability and user satisfaction. However, existing QoS prediction methods often fail to capture rich contextual information and exhibit poor performance under extreme data sparsity and structural noise. To bridge this gap, we propose a novel architecture, QoSMGAA, specifically designed to enhance prediction accuracy in complex and noisy network service environments. QoSMGAA integrates a multi-order attention mechanism to aggregate extensive contextual data and predict missing QoS values effectively. Additionally, our method incorporates adversarial neural networks to perform autoregressive supervised learning based on transformed interaction matrices. To capture complex, higher-order interactions among users and services, we employ a discrete sampling technique leveraging the Gumbel-Softmax method to generate informative negative samples. Comprehensive experimental validation conducted on large-scale real-world datasets demonstrates that our proposed model significantly outperforms existing baseline methods, highlighting its strong potential for practical deployment in service selection and recommendation scenarios.

Figures (9)

• Figure 1: An illustration of user-service interactions and associated QoS parameters
• Figure 2: Framework of QoSMGAA.
• Figure 3: Ablation study on MOGAT
• Figure 4: Ablation study on Adversarial Neural Networks
• Figure 5: Impact of Attention Orders in the Multi-Order Graph Attention Mechanism