ImbalancE: Inference-Time Latent Search Against Degree Imbalance in Link Prediction

2026-09-29

AI Summary

This study addresses the significant performance degradation in knowledge graph link prediction caused by entity degree imbalance, particularly for low-degree (long-tail) entities. To mitigate this issue, it proposes an inference-time latent search optimization framework that introduces a novel dynamic search mechanism based on pretrained models. By exploring and integrating multi-source information within the embedding space, this approach effectively corrects bias without requiring model retraining. Furthermore, this work reveals the pervasive impact of degree imbalance on prediction quality. Experimental results demonstrate that the proposed method substantially outperforms conventional approaches on imbalanced triplets across benchmark datasets, significantly enhancing both prediction accuracy and model usability in long-tail scenarios.

Abstract

Knowledge Graph Embedding models have been extensively used to learn representations of entities and relations in Knowledge Graphs for predicting missing links. However, the quality of the learned representations varies a lot across different areas of the graph. If previous research has loosely linked the problem to relation types or degree bias, we show that it is more widespread and it correlates with the degree imbalance of the entities in test triples. In particular, the prediction of a target entity that has a degree much smaller than the degree of the anchor entity is extremely problematic. This is critical in recommender systems and other use cases, where these triples represent important corner cases. To address this issue, we propose an inference-time latent search optimization method capable of significantly improving model predictions on the most imbalanced triples. Built on top of a pre-trained model, it explores the embedding space at evaluation time, blending known and out-of-band information to mitigate the degree imbalance bias. We show the value of our approach on imbalanced triples from common benchmark datasets, where we outperform conventional methods, opening the door to the successful adoption of Knowledge Graph Embedding models on these critical corner cases.

Problem

• Knowledge Graph Embedding
• Link Prediction
• Degree Imbalance
• Representation Quality

Innovation

• Inference-Time Optimization
• Latent Search
• Degree Imbalance
• Link Prediction
• Knowledge Graph Embedding