# Sampling-Decomposable Generative Adversarial Recommender

arXiv: https://arxiv.org/abs/2011.00956 · HTML: https://mlsanity.com/paper/2011.00956/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Nov 2020.

## Built on by (most cited first, 16 of 16)

- [Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference Optimization](https://mlsanity.com/paper/2407.07880/) (2024)
- [Denoising Diffusion Recommender Model](https://mlsanity.com/paper/2401.06982/) (2024)
- [Understanding Contrastive Learning via Distributionally Robust Optimization](https://mlsanity.com/paper/2310.11048/) (2023)
- [RecDiff: Diffusion Model for Social Recommendation](https://mlsanity.com/paper/2406.01629/) (2024)
- [Learning from Heterogeneity: Generalizing Dynamic Facial Expression Recognition via Distributionally Robust Optimization](https://mlsanity.com/paper/2507.15765/) (2025)
- [Does Negative Sampling Matter? A Review with Insights into its Theory and Applications](https://mlsanity.com/paper/2402.17238/) (2024)
- [Hyperbolic Diffusion Recommender Model](https://mlsanity.com/paper/2504.01541/) (2025)
- [Fast Variational AutoEncoder with Inverted Multi-Index for Collaborative Filtering](https://mlsanity.com/paper/2109.05773/) (2021)
- [Progressively Optimized Bi-Granular Document Representation for Scalable Embedding Based Retrieval](https://mlsanity.com/paper/2201.05409/) (2022)
- [Cooperative Retriever and Ranker in Deep Recommenders](https://mlsanity.com/paper/2206.14649/) (2022)
- [Generative Diffusion Models for Sequential Recommendations](https://mlsanity.com/paper/2410.19429/) (2024)
- [Towards Robust Recommender Systems via Triple Cooperative Defense](https://mlsanity.com/paper/2210.13762/) (2022)
- [When and What to Recommend: Joint Modeling of Timing and Content for Active Sequential Recommendation](https://mlsanity.com/paper/2511.18717/) (2025)
- [A generative recommender system with GMM prior for cancer drug generation and sensitivity prediction](https://mlsanity.com/paper/2206.03555/) (2022)
- [Securing Recommender System via Cooperative Training](https://mlsanity.com/paper/2401.12700/) (2024)
- [Multi-Margin Cosine Loss: Proposal and Application in Recommender Systems](https://mlsanity.com/paper/2405.04614/) (2024)

## Abstract (arXiv)

Recommendation techniques are important approaches for alleviating information overload. Being often trained on implicit user feedback, many recommenders suffer from the sparsity challenge due to the lack of explicitly negative samples. The GAN-style recommenders (i.e., IRGAN) addresses the challenge by learning a generator and a discriminator adversarially, such that the generator produces increasingly difficult samples for the discriminator to accelerate optimizing the discrimination objective. However, producing samples from the generator is very time-consuming, and our empirical study shows that the discriminator performs poor in top-k item recommendation. To this end, a theoretical analysis is made for the GAN-style algorithms, showing that the generator of limit capacity is diverged from the optimal generator. This may interpret the limitation of discriminator's performance. Based on these findings, we propose a Sampling-Decomposable Generative Adversarial Recommender (SD-GAR). In the framework, the divergence between some generator and the optimum is compensated by self-normalized importance sampling; the efficiency of sample generation is improved with a sampling-decomposable generator, such that each sample can be generated in O(1) with the Vose-Alias method. Interestingly, due to decomposability of sampling, the generator can be optimized with the closed-form solutions in an alternating manner, being different from policy gradient in the GAN-style algorithms. We extensively evaluate the proposed algorithm with five real-world recommendation datasets. The results show that SD-GAR outperforms IRGAN by 12.4% and the SOTA recommender by 10% on average. Moreover, discriminator training can be 20x faster on the dataset with more than 120K items.