LGFeb 10, 2023

Toward Degree Bias in Embedding-Based Knowledge Graph Completion

arXiv:2302.05044v10.3231 citationsh-index: 90
AI Analysis70

This addresses bias issues in knowledge graph completion for AI applications, representing an incremental improvement with a novel method for a known bottleneck.

The paper tackles the problem of degree bias in embedding-based knowledge graph completion, validating its existence and identifying key factors, and introduces KG-Mixup, a data augmentation method that improves various KGC methods and outperforms others on benchmark datasets.

A fundamental task for knowledge graphs (KGs) is knowledge graph completion (KGC). It aims to predict unseen edges by learning representations for all the entities and relations in a KG. A common concern when learning representations on traditional graphs is degree bias. It can affect graph algorithms by learning poor representations for lower-degree nodes, often leading to low performance on such nodes. However, there has been limited research on whether there exists degree bias for embedding-based KGC and how such bias affects the performance of KGC. In this paper, we validate the existence of degree bias in embedding-based KGC and identify the key factor to degree bias. We then introduce a novel data augmentation method, KG-Mixup, to generate synthetic triples to mitigate such bias. Extensive experiments have demonstrated that our method can improve various embedding-based KGC methods and outperform other methods tackling the bias problem on multiple benchmark datasets.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes