# Hypergraph Semi-Supervised Contrastive Learning for Hyperedge Prediction Based on Enhanced Attention Aggregator

## TL;DR

This paper introduces a new method for predicting hyperedges in complex systems by using enhanced attention and contrastive learning techniques.

## Contribution

The novel contribution is OFSH, a hyperedge prediction framework that incorporates order propagation and contrastive learning to handle data sparsity and node heterogeneity.

## Key findings

- OFSH significantly improves hyperedge prediction performance on real-world datasets.
- The method effectively captures high-order semantics through key node-guided augmentation and adaptive masking.
- Triadic contrastive loss enhances cross-view consistency and semantic invariance.

## Abstract

Hyperedge prediction is crucial for uncovering higher-order relationships in complex systems but faces core challenges, including unmodeled node influence heterogeneity, overlooked hyperedge order effects, and data sparsity. This paper proposes Order propagation Fusion Self-supervised learning for Hyperedge prediction (OFSH) to address these issues. OFSH introduces a hyperedge order propagation mechanism that dynamically learns node importance weights and groups neighbor hyperedges by order, applying max–min pooling to amplify feature distinctions. To mitigate data sparsity, OFSH incorporates a key node-guided augmentation strategy with adaptive masking, preserving core high-order semantics. It identifies topological hub nodes based on their comprehensive influence and employs adaptive masking probabilities to generate augmented views preserving core high-order semantics. Finally, a triadic contrastive loss is employed to maximize cross-view consistency and capture invariant semantic information under perturbations. Extensive experiments on five public real-world hypergraph datasets demonstrate significant improvements over state-of-the-art methods in AUROC and AP.