Mind the Links: Cross-Layer Attention for Link Prediction in Multiplex Networks

2025-09-27

AI Summary

To address insufficient cross-layer dependency modeling and poor scalability in multilayer network link prediction, this paper reformulates the task as a multi-view edge classification problem and proposes the Cross-Layer Self-Attention (CLSA) framework. CLSA constructs edge-view sequences per layer and employs self-attention to dynamically integrate cross-layer structural evidence. It introduces two compatible variants—Trans-SLE (leveraging static embeddings) and Trans-GAT (integrating GNN encoders)—and adopts a leakage-free evaluation protocol with a Union-Set candidate pool to ensure fairness and computational efficiency. Extensive experiments on six public multilayer network datasets demonstrate that CLSA consistently outperforms state-of-the-art baselines—including MELL, HOPLP-MUL, and RMNE—with average macro-F₁ gains of 3.2–9.7 percentage points. The results validate CLSA’s effectiveness, generalizability across diverse network topologies, and scalability to large-scale multilayer graphs.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Natural Language Processing: Machine Translation, Multilinguality, Cross-Lingual NLP
• Computer Vision: Multi-modal Vision

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search

Abstract

Multiplex graphs capture diverse relations among shared nodes. Most predictors either collapse layers or treat them independently. This loses crucial inter-layer dependencies and struggles with scalability. To overcome this, we frame multiplex link prediction as multi-view edge classification. For each node pair, we construct a sequence of per-layer edge views and apply cross-layer self-attention to fuse evidence for the target layer. We present two models as instances of this framework: Trans-SLE, a lightweight transformer over static embeddings, and Trans-GAT, which combines layer-specific GAT encoders with transformer fusion. To ensure scalability and fairness, we introduce a Union--Set candidate pool and two leakage-free protocols: cross-layer and inductive subgraph generalization. Experiments on six public multiplex datasets show consistent macro-F_1 gains over strong baselines (MELL, HOPLP-MUL, RMNE). Our approach is simple, scalable, and compatible with both precomputed embeddings and GNN encoders.

Problem

• Predicting links in multiplex networks using cross-layer attention
• Addressing inter-layer dependency loss in existing prediction methods
• Ensuring scalability and fairness with leakage-free generalization protocols

Innovation

• Uses cross-layer self-attention to fuse multiplex edge evidence
• Combines GAT encoders with transformer fusion for layers
• Introduces leakage-free protocols for scalable generalization