MM-GEF: Multi-modal representation meet collaborative filtering

TL;DR

MM-GEF introduces a graph-based early-fusion approach that combines multi-modal item features with collaborative signals, significantly enhancing recommendation accuracy in e-commerce settings.

Contribution

The paper presents a novel graph-based early-fusion method that integrates multi-modal content features with collaborative signals for improved recommendations.

Findings

Significant performance improvements over state-of-the-art methods

Effective integration of multi-modal features with collaborative signals

Robust results across four public datasets

Abstract

In modern e-commerce, item content features in various modalities offer accurate yet comprehensive information to recommender systems. The majority of previous work either focuses on learning effective item representation during modelling user-item interactions, or exploring item-item relationships by analysing multi-modal features. Those methods, however, fail to incorporate the collaborative item-user-item relationships into the multi-modal feature-based item structure. In this work, we propose a graph-based item structure enhancement method MM-GEF: Multi-Modal recommendation with Graph Early-Fusion, which effectively combines the latent item structure underlying multi-modal contents with the collaborative signals. Instead of processing the content feature in different modalities separately, we show that the early-fusion of multi-modal features provides significant improvement. MM-GEF learns refined item representations by injecting structural information obtained from both multi-modal and collaborative signals. Through extensive experiments on four publicly available datasets, we demonstrate systematical improvements of our method over state-of-the-art multi-modal recommendation methods.