HyGEN: Regularizing Negative Hyperedge Generation for Accurate Hyperedge Prediction

2025-02-09

AI Summary

To address the challenge of unguided negative sampling and spurious negative hyperedge generation caused by data sparsity in higher-order relational prediction, this paper proposes a regularized, positive-guided negative hyperedge generation framework. Methodologically, it integrates hypergraph neural embedding with differentiable negative sampling for end-to-end training. Its key contributions are: (1) the first differentiable negative hyperedge generator, which leverages positive hyperedges as structural priors to produce semantically meaningful negatives; and (2) a false-negative-aware regularization term that jointly optimizes the generator and discriminator to explicitly suppress mislabeling risk. Evaluated on six real-world hypergraph datasets, the method consistently outperforms four state-of-the-art baselines, achieving average AUC improvements of 3.2–7.8 percentage points. These results demonstrate its effectiveness in mitigating data sparsity and enhancing negative sampling quality.

Technology Category

Machine Learning: Graph-based Machine LearningData Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB CompletionReasoning under Uncertainty: Relational Probabilistic Models

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 graphsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Hyperedge prediction is a fundamental task to predict future high-order relations based on the observed network structure. Existing hyperedge prediction methods, however, suffer from the data sparsity problem. To alleviate this problem, negative sampling methods can be used, which leverage non-existing hyperedges as contrastive information for model training. However, the following important challenges have been rarely studied: (C1) lack of guidance for generating negatives and (C2) possibility of producing false negatives. To address them, we propose a novel hyperedge prediction method, HyGEN, that employs (1) a negative hyperedge generator that employs positive hyperedges as a guidance to generate more realistic ones and (2) a regularization term that prevents the generated hyperedges from being false negatives. Extensive experiments on six real-world hypergraphs reveal that HyGEN consistently outperforms four state-of-the-art hyperedge prediction methods.

Problem

• Predicts future high-order network relations
• Addresses data sparsity in hyperedge prediction
• Prevents generation of false negative hyperedges

Innovation

• Negative hyperedge generator
• Regularization term
• Positive hyperedge guidance