Progressively Modality Freezing for Multi-Modal Entity Alignment

TL;DR

This paper introduces a progressive modality freezing strategy for multi-modal entity alignment, improving feature relevance and modal consistency, leading to state-of-the-art results across multiple datasets.

Contribution

It proposes a novel PMF approach with a cross-modal association loss to enhance multi-modal feature fusion and alignment accuracy.

Findings

Achieves state-of-the-art performance on nine datasets.

Effectively enhances modal consistency and feature relevance.

Demonstrates the benefits of freezing modalities during training.

Abstract

Multi-Modal Entity Alignment aims to discover identical entities across heterogeneous knowledge graphs. While recent studies have delved into fusion paradigms to represent entities holistically, the elimination of features irrelevant to alignment and modal inconsistencies is overlooked, which are caused by inherent differences in multi-modal features. To address these challenges, we propose a novel strategy of progressive modality freezing, called PMF, that focuses on alignmentrelevant features and enhances multi-modal feature fusion. Notably, our approach introduces a pioneering cross-modal association loss to foster modal consistency. Empirical evaluations across nine datasets confirm PMF's superiority, demonstrating stateof-the-art performance and the rationale for freezing modalities. Our code is available at https://github.com/ninibymilk/PMF-MMEA.