Do We Really Need to Drop Items with Missing Modalities in Multimodal Recommendation?

TL;DR

This paper challenges the common practice of dropping items with missing modalities in multimodal recommendation systems, proposing instead a novel imputation pipeline that improves recommendation performance by leveraging graph structures and similarity measures.

Contribution

It introduces a new imputation pipeline for missing modalities in multimodal recommendation, utilizing graph-based and similarity-based methods, which outperforms data filtering approaches.

Findings

Imputation improves recommendation accuracy over dropping missing data.

Graph and similarity-based imputation methods outperform traditional strategies.

Pre-processing with imputation is beneficial and should replace data filtering.

Abstract

Generally, items with missing modalities are dropped in multimodal recommendation. However, with this work, we question this procedure, highlighting that it would further damage the pipeline of any multimodal recommender system. First, we show that the lack of (some) modalities is, in fact, a widely-diffused phenomenon in multimodal recommendation. Second, we propose a pipeline that imputes missing multimodal features in recommendation by leveraging traditional imputation strategies in machine learning. Then, given the graph structure of the recommendation data, we also propose three more effective imputation solutions that leverage the item-item co-purchase graph and the multimodal similarities of co-interacted items. Our method can be plugged into any multimodal RSs in the literature working as an untrained pre-processing phase, showing (through extensive experiments) that any data pre-filtering is not only unnecessary but also harmful to the performance.