Context-Driven Knowledge Graph Completion with Semantic-Aware Relational Message Passing

2025-06-29

AI Summary

Traditional message-passing mechanisms in knowledge graph completion (KGC) suffer from noise injection, information dilution, and over-smoothing due to indiscriminate aggregation of neighboring edges. To address this, we propose a semantic-aware relational message-passing framework. Our method introduces two key innovations: (1) an edge-level semantic similarity measure defined in a shared latent space, enabling semantic-driven Top-K neighbor selection; and (2) a multi-head attention aggregator that models context-sensitive message propagation within triplets. By explicitly incorporating relational semantics into neighborhood aggregation, the framework significantly enhances the semantic focus of node representations. Experimental results demonstrate state-of-the-art performance on standard benchmarks—including FB15k-237 and WN18RR—outperforming both GNN-based and translation-based KGC approaches. The framework effectively suppresses interference from irrelevant edges and improves link prediction accuracy.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Natural Language Processing: Sentence-level Semantics, Textual Inference, etc.
• Machine Learning: Graph-based Machine Learning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Semantic context surrounding a triplet $(h, r, t)$ is crucial for Knowledge Graph Completion (KGC), providing vital cues for prediction. However, traditional node-based message passing mechanisms, when applied to knowledge graphs, often introduce noise and suffer from information dilution or over-smoothing by indiscriminately aggregating information from all neighboring edges. To address this challenge, we propose a semantic-aware relational message passing. A core innovation of this framework is the introduction of a extbf{semantic-aware Top-K neighbor selection strategy}. Specifically, this strategy first evaluates the semantic relevance between a central node and its incident edges within a shared latent space, selecting only the Top-K most pertinent ones. Subsequently, information from these selected edges is effectively fused with the central node's own representation using a extbf{multi-head attention aggregator} to generate a semantically focused node message. In this manner, our model not only leverages the structure and features of edges within the knowledge graph but also more accurately captures and propagates the contextual information most relevant to the specific link prediction task, thereby effectively mitigating interference from irrelevant information. Extensive experiments demonstrate that our method achieves superior performance compared to existing approaches on several established benchmarks.

Problem

• Improves Knowledge Graph Completion by leveraging semantic context
• Reduces noise and information dilution in message passing
• Selects Top-K relevant edges for accurate link prediction

Innovation

• Semantic-aware Top-K neighbor selection strategy
• Multi-head attention aggregator for fusion
• Context-driven knowledge graph completion