Mining Stable Preferences: Adaptive Modality Decorrelation for Multimedia Recommendation
Jinghao Zhang, Qiang Liu, Shu Lei Wu, Liang Wang · 2023
Multimedia content is of predominance in the modern Web era. Many recommender models have been proposed to investigate how users interact with items which are represented in diverse modalities. In real scenarios, different modalities reveal different aspects of item attributes and usually possess different importance to user purchase decisions. However, it is difficult for models to figure out users' true preference towards different modalities since there exists strong statistical correlation between different modalities. Even worse, the strong statistical correlation might mislead models to learn the spurious preference towards inconsequential modalities. As a result, when data (modal features) distribution shifts, the learned spurious preference might not guarantee to be as effective on inference as on the training set.