Mitigating Spurious Correlations for Self-supervised Recommendation

TL;DR

This work addresses spurious correlations in self-supervised recommendation by introducing Invariant Feature Learning (IFL), which automatically masks spurious features and blocks their negative transmission through mask-guided contrastive learning. By clustering interactions into multiple environments and enforcing invariant features via a gradient-variance regularizer, IFL identifies robust representations while discarding misleading cues. The method optimizes a composite objective $\mathcal{L}=\mathcal{L}_{CF}+\alpha\mathcal{L}_{ssl}+\beta\mathcal{L}_v+\lambda\|\boldsymbol{\theta}\|^2$, and uses mask-driven data augmentation to stabilize SSL against spurious features. Empirical results on Meituan and XING show improved OOD generalization without sacrificing IID performance, demonstrating the practical potential of invariant feature masking in SSL-based recommendations, with code available at the provided GitHub repository.

Abstract

Recent years have witnessed the great success of self-supervised learning (SSL) in recommendation systems. However, SSL recommender models are likely to suffer from spurious correlations, leading to poor generalization. To mitigate spurious correlations, existing work usually pursues ID-based SSL recommendation or utilizes feature engineering to identify spurious features. Nevertheless, ID-based SSL approaches sacrifice the positive impact of invariant features, while feature engineering methods require high-cost human labeling. To address the problems, we aim to automatically mitigate the effect of spurious correlations. This objective requires to 1) automatically mask spurious features without supervision, and 2) block the negative effect transmission from spurious features to other features during SSL. To handle the two challenges, we propose an invariant feature learning framework, which first divides user-item interactions into multiple environments with distribution shifts and then learns a feature mask mechanism to capture invariant features across environments. Based on the mask mechanism, we can remove the spurious features for robust predictions and block the negative effect transmission via mask-guided feature augmentation. Extensive experiments on two datasets demonstrate the effectiveness of the proposed framework in mitigating spurious correlations and improving the generalization abilities of SSL models. The code is available at https://github.com/Linxyhaha/IFL.

Figures (14)

• Figure 1: An example of spurious correlations in job recommendation, where the positive interactions with the full-time jobs show strong correlations with the users' 4 and 6 years of experience.
• Figure 2: Illustration of IFL framework. The two-tower encoders on the left are used for normal recommender training with the CF loss and variance loss. On the right side, we do mask-guided feature augmentation over the user and item features, i.e., drop the spurious features and then conduct contrastive learning. The augmentations on users and items are the same and we only show that on items to save space.
• Figure 3: Illustration of the environment division in feature mask learning. The interaction representations $\bm{z}_{ui}$ are clustered into $C=2$ environments. The distributions of spurious features shift across two environments while those of invariant features are stable.
• Figure 4: Illustration of mask-guided contrastive learning in IFL. (a) shows the augmentation by dropping spurious features, (b) demonstrates the contrastive pairs in a sample batch, and (c) presents the contrastive loss between two items. Only two samples' contrastive pairs are shown in (b) for neatness.
• Figure 5: The conditional distribution of the interaction given spurious features for the IID and OOD sets of XING, where the spurious feature is the user work experience and the interaction is about interacting with the full-time job.