# Re4: Learning to Re-contrast, Re-attend, Re-construct for Multi-interest Recommendation

arXiv: https://arxiv.org/abs/2208.08011 · HTML: https://mlsanity.com/paper/2208.08011/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Aug 2022.

## Built on by (most cited first, 19 of 19)

- [DiffuRec: A Diffusion Model for Sequential Recommendation](https://mlsanity.com/paper/2304.00686/) (2023)
- [MISSRec: Pre-training and Transferring Multi-modal Interest-aware Sequence Representation for Recommendation](https://mlsanity.com/paper/2308.11175/) (2023)
- [Disentangled Contrastive Collaborative Filtering](https://mlsanity.com/paper/2305.02759/) (2023)
- [When Search Meets Recommendation: Learning Disentangled Search Representation for Recommendation](https://mlsanity.com/paper/2305.10822/) (2023)
- [Facet-Aware Multi-Head Mixture-of-Experts Model for Sequential Recommendation](https://mlsanity.com/paper/2411.01457/) (2024)
- [DGenCTR: Towards a Universal Generative Paradigm for Click-Through Rate Prediction via Discrete Diffusion](https://mlsanity.com/paper/2508.14500/) (2025)
- [Group-Aware Interest Disentangled Dual-Training for Personalized Recommendation](https://mlsanity.com/paper/2311.09577/) (2023)
- [BiVRec: Bidirectional View-based Multimodal Sequential Recommendation](https://mlsanity.com/paper/2402.17334/) (2024)
- [Long-term User Engagement Optimization through Model-agnostic Downstream Rewards Learning](https://mlsanity.com/paper/2607.14192/) (2026)
- [Graph Exploration Matters: Improving both individual-level and system-level diversity in WeChat Feed Recommender](https://mlsanity.com/paper/2306.00009/) (2023)
- [Multi-Tower Multi-Interest Recommendation with User Representation Repel](https://mlsanity.com/paper/2403.05122/) (2024)
- [Orthogonal Hyper-category Guided Multi-interest Elicitation for Micro-video Matching](https://mlsanity.com/paper/2407.14741/) (2024)
- [CoActionGraphRec: Sequential Multi-Interest Recommendations Using Co-Action Graphs](https://mlsanity.com/paper/2410.11464/) (2024)
- [Burger: Robust Graph Denoising-augmentation Fusion and Multi-semantic Modeling in Social Recommendation](https://mlsanity.com/paper/2505.06612/) (2025)
- [MATT-CTR: Unleashing a Model-Agnostic Test-Time Paradigm for CTR Prediction with Confidence-Guided Inference Paths](https://mlsanity.com/paper/2510.08932/) (2025)
- [GemiRec: Interest Quantization and Generation for Multi-Interest Recommendation](https://mlsanity.com/paper/2510.14626/) (2025)
- [MergeRec: Model Merging for Data-Isolated Cross-Domain Sequential Recommendation](https://mlsanity.com/paper/2601.01753/) (2026)
- [Facet-Aware Multi-Head Mixture-of-Experts Model with Text-Enhanced Pre-training for Sequential Recommendation](https://mlsanity.com/paper/2601.12301/) (2026)
- [TME-PSR: Time-aware, Multi-interest, and Explanation Personalization for Sequential Recommendation](https://mlsanity.com/paper/2604.09439/) (2026)

## Abstract (arXiv)

Effectively representing users lie at the core of modern recommender systems. Since users' interests naturally exhibit multiple aspects, it is of increasing interest to develop multi-interest frameworks for recommendation, rather than represent each user with an overall embedding. Despite their effectiveness, existing methods solely exploit the encoder (the forward flow) to represent multiple aspects of interests. However, without explicit regularization, the interest embeddings may not be distinct from each other nor semantically reflect representative historical items. Towards this end, we propose the Re4 framework, which leverages the backward flow to reexamine each interest embedding. Specifically, Re4 encapsulates three backward flows, i.e., 1) Re-contrast, which drives each interest embedding to be distinct from other interests using contrastive learning; 2) Re-attend, which ensures the interest-item correlation estimation in the forward flow to be consistent with the criterion used in final recommendation; and 3) Re-construct, which ensures that each interest embedding can semantically reflect the information of representative items that relate to the corresponding interest. We demonstrate the novel forward-backward multi-interest paradigm on ComiRec, and perform extensive experiments on three real-world datasets. Empirical studies validate that Re4 helps to learn learning distinct and effective multi-interest representations.