Efficient Recommendations via Graph Coarsening and Label Propagation

2026-07-24

AI Summary

This work addresses the challenge of balancing efficiency and accuracy in large-scale graph-based recommendation systems. The authors propose a two-stage diffusion framework that first coarsens the graph using domain-informed heuristics to preserve critical business relationships, then performs label propagation on the coarsened graph to generate coarse-grained predictions. A subsequent local refinement step propagates labels within user-specific subgraphs to produce fine-grained, user-level recommendations. By integrating graph coarsening, multi-step label propagation, and a lightweight graph neural network (GNN), the method achieves substantial performance gains over full-graph baselines on real-world telecom datasets: it improves NDCG@5 by 24% over a pure label propagation approach, and yields over 50% further improvement when augmented with the lightweight GNN, while simultaneously maintaining scalability, low latency, and high accuracy.

Abstract

Graph-based recommendations are widely adopted in real-world industrial applications. However, graphs in these systems often reach a massive scale, posing notable scalability and efficiency challenges. This requires techniques that can effectively balance predictive quality with computational cost. One promising approach is graph coarsening, an adaptive graph reduction technique that offers a way to systematically construct smaller, yet structurally representative, versions of the original large-scale graphs. In this work, we propose a flexible two-stage diffusion framework that combines graph coarsening with multi-step label propagation in the telecommunications domain. Domain-specific heuristics are applied to first aggregate nodes into meaningful communities, reducing graph size while preserving essential business-relevant relationships. An initial diffusion process done by a Label Propagation Algorithm (LPA) or a Graph Neural Network (GNN) propagates labels across the coarsened graph to produce coarse-grained predictions. Finally, a second LPA within subgraphs generates the final recommendations for individual users. On a real-world telecommunications dataset, when using LPA in both stages, our method achieves up to +24% NDCG@5 over the full-graph LPA baseline. Incorporating a lightweight GNN in the first stage further boosts NDCG@5 by more than 50%, but requires substantial training and inference time. Through extensive experiments and a detailed ablation, we quantify these trade-offs and demonstrate that our coarsening-driven approach delivers an optimal balance between scalability, latency, and recommendation quality.

Problem

• graph-based recommendations
• scalability
• efficiency
• graph coarsening
• label propagation

Innovation

• graph coarsening
• label propagation
• two-stage diffusion
• scalable recommendation
• telecommunications