# M^2VAE: Multi-Modal Multi-View Variational Autoencoder for Cold-start Item Recommendation

arXiv: https://arxiv.org/abs/2508.00452 · HTML: https://mlsanity.com/paper/2508.00452/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in MoE for recommender systems, one of 12 directions in Mixture-of-experts.
- Builds on 19 of its 41 references that are in the map.
- First posted on arXiv 1 Aug 2025.

## Placed in

- Mixture-of-experts / [MoE for recommender systems](https://mlsanity.com/moe/moe-for-recommender-systems/)

## Builds on (most cited first, 19 of 19)

- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [Auto-Encoding Variational Bayes](https://mlsanity.com/paper/1312.6114/) (2013)
- [Tutorial on Variational Autoencoders](https://mlsanity.com/paper/1606.05908/) (2016)
- [Variational Autoencoders for Collaborative Filtering](https://mlsanity.com/paper/1802.05814/) (2018)
- [Neural Factorization Machines for Sparse Predictive Analytics](https://mlsanity.com/paper/1708.05027/) (2017)
- [Variational Mixture-of-Experts Autoencoders for Multi-Modal Deep Generative Models](https://mlsanity.com/paper/1911.03393/) (2019)
- [Variational Autoencoder for Deep Learning of Images, Labels and Captions](https://mlsanity.com/paper/1609.08976/) (2016)
- [MeLU: Meta-Learned User Preference Estimator for Cold-Start Recommendation](https://mlsanity.com/paper/1908.00413/) (2019)
- [Joint Multimodal Learning with Deep Generative Models](https://mlsanity.com/paper/1611.01891/) (2016)
- [Contrastive Learning for Cold-Start Recommendation](https://mlsanity.com/paper/2107.05315/) (2021)
- [RecVAE: a New Variational Autoencoder for Top-N Recommendations with Implicit Feedback](https://mlsanity.com/paper/1912.11160/) (2019)
- [Multi-VAE: Learning Disentangled View-common and View-peculiar Visual Representations for Multi-view Clustering](https://mlsanity.com/paper/2106.11232/) (2021)
- [Learning to Warm Up Cold Item Embeddings for Cold-start Recommendation with Meta Scaling and Shifting Networks](https://mlsanity.com/paper/2105.04790/) (2021)
- [Contrastive Collaborative Filtering for Cold-Start Item Recommendation](https://mlsanity.com/paper/2302.02151/) (2023)
- [Learning Graph Meta Embeddings for Cold-Start Ads in Click-Through Rate Prediction](https://mlsanity.com/paper/2105.08909/) (2021)
- [Mutually-Regularized Dual Collaborative Variational Auto-encoder for Recommendation Systems](https://mlsanity.com/paper/2211.11662/) (2022)
- [Auditory Attention Decoding with Task-Related Multi-View Contrastive Learning](https://mlsanity.com/paper/2308.04244/) (2023)
- [Warming Up Cold-Start CTR Prediction by Learning Item-Specific Feature Interactions](https://mlsanity.com/paper/2407.10112/) (2024)
- [Firzen: Firing Strict Cold-Start Items with Frozen Heterogeneous and Homogeneous Graphs for Recommendation](https://mlsanity.com/paper/2410.07654/) (2024)

## Built on by (most cited first, 1 of 1)

- [Simple Yet Effective Selective Imputation for Incomplete Multi-view Clustering](https://mlsanity.com/paper/2512.10327/) (2025)

## Abstract (arXiv)

Cold-start item recommendation is a significant challenge in recommendation systems, particularly when new items are introduced without any historical interaction data. While existing methods leverage multi-modal content to alleviate the cold-start issue, they often neglect the inherent multi-view structure of modalities, the distinction between shared and modality-specific features. In this paper, we propose Multi-Modal Multi-View Variational AutoEncoder (M^2VAE), a generative model that addresses the challenges of modeling common and unique views in attribute and multi-modal features, as well as user preferences over single-typed item features. Specifically, we generate type-specific latent variables for item IDs, categorical attributes, and image features, and use Product-of-Experts (PoE) to derive a common representation. A disentangled contrastive loss decouples the common view from unique views while preserving feature informativeness. To model user inclinations, we employ a preference-guided Mixture-of-Experts (MoE) to adaptively fuse representations. We further incorporate co-occurrence signals via contrastive learning, eliminating the need for pretraining. Extensive experiments on real-world datasets validate the effectiveness of our approach.