基于扩散模型的生成对抗式推荐方法

doi:10.19678/j.issn.1000-3428. 0253386

摘要/Abstract

摘要： 生成式模型因其有效的数据生成能力，近年来被广泛应用于推荐系统领域。然而，现有生成式推荐系统由于模型生成过程的随机性导致稳定性不足，并且有限的表示学习能力影响了个性化推荐的准确性。为了解决上述问题，提出一种基于扩散模型的生成对抗式推荐方法。具体来说，首先通过变分自编码器 (VAE) 对原始向量进行压缩，然后使用扩散模型在隐式空间进行多步加噪和去噪，学习高质量用户表示。此外，引入对抗训练机制为去噪过程提出反馈信号，缓解其生成过程不可控的问题。在Amazon-book、Yelp和Movielens-1M三个公开数据集上进行实验，所提出的方法相比主要基线在召回率 (Recall@10)和归一化折损累计增益 (NDCG@10)上分别最高提升20.3%、18.9%，说明方法能够有效预测用户行为，提高推荐精度。

Abstract: Generative models have achieved remarkable results due to their effective data generation capabilities and have been widely applied in the field of recommendation systems in recent years. Generative recommendation, through probabilistic modeling, directly learns the potential distribution of users' historical behaviors and generates possible interaction scenarios, breaking through the traditional retrieval paradigm and becoming a research hotspot in the field of recommendation systems. However, the existing generative recommendation systems have insufficient stability due to the randomness of the model generation process, and the limited representation learning ability affects the accuracy of personalized recommendations. To solve the above problems, a generative adversarial recommendation method based on diffusion model is proposed. Specifically, first of all, to alleviate the resource consumption caused by direct diffusion, the original vector is compressed through a Variational Autoencoder (VAE). Then, the diffusion model is used in latent space for multi-step denoising noising and denoising to learn high-quality user representations. In addition, an adversarial training mechanism is introduced to provide feedback signals for the denoising process, alleviating the problem of uncontrollability in its generation process. Experiments were conducted on three public datasets, namely Amazon-book, Yelp and Movielens-1M. Compared with the baseline model, the proposed method has significant improvements in the metrics of Recall and normalized cumulative loss gain (NDCG), indicating that the method can effectively predict user behavior. Improve the accuracy of recommendations.

朱奕健, 冒睿瑞. 基于扩散模型的生成对抗式推荐方法[J]. 计算机工程, doi: 10.19678/j.issn.1000-3428. 0253386.

ZHU Yijian, MAO Ruirui. Generative Adversarial Recommendation Method Based on Diffusion Model[J]. Computer Engineering, doi: 10.19678/j.issn.1000-3428. 0253386.

参考文献

[1] 余正涛,孙资钦,张勇丙,等.基于随机自注意力和动量对比学习的自监督序列推荐方法[J/OL].计算机工程: 1-16[2025-07-30].https://doi.org/10.19678/j.issn.1000-3428.0070193. Yu Z T, Sun Z Q, Zhang Y R, et al. Self-supervised sequence recommendation method based on random self-attention and momentum contrastive learning[J/OL]. Computer Engineering: 1-16[2025-07-30]. https://doi.org/10.19678/j.issn.1000-3428.0070193. (in Chinese)
[2] Deldjoo Y, He Z K, McAuley J, et al. A Review of Modern Recommender Systems Using Generative Models (Gen-RecSys)[C]//Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York, USA: Association for Computing Machinery, 2024: 6448-6458.
[3] Chen H, Bei Y C, Shen Q J, et al. Macro Graph Neural Networks for Online Billion-Scale Recommender Systems[C]//Proceedings of the ACM Web Conference 2024. New York, USA: Association for Computing Machinery, 2024: 3598-3608.
[4] Li Y Q, Lin X Y, Wang W J, et al. A Survey of Generative Search and Recommendation in the Era of L-arge Language Models[DB/OL]. arXiv, 2024[2025-07-30]. https://arxiv.org/abs/2404.16924.
[5] Deldjoo Y, Mehta N, Sathiamoorthy M, et al. Toward Holistic Evaluation of Recommender Systems Powered by Generative Models[C]//Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: Association for Computing Machinery, 2025: 3932-3942.
[6] Gui J, Sun Z N, Wen Y G, et al. A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications[J]. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(4): 3313-3332.
[7] Liang D, Krishnan R G., Hoffman M D, et al. Variational Autoencoders for Collaborative Filtering[C]//Proceedings of the 2018 World Wide Web Conference. Republic and Canton of Geneva, CHE: International World Wide Web Conferences Steering Committee, 2018: 689-698.
[8] Wang J, Yu L T, Zhang W N, et al. IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models[C]//Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: Association for Computing Machinery, 2017: 515-524.
[9] Xu Y Y, Wang W J, Feng F L, et al. Diffusion Models for Generative Outfit Recommendation[C]//Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: Association for Computing Machinery, 2024: 1350-1359.
[10] Bao K Q, Zhang J Z, Zhang Y, et al. TALLRec: An Effective and Efficient Tuning Framework to Align Large Language Model with Recommendation[C]//Proceedings of the 17th ACM Conference on Recommender Systems. New York, USA: Association for Computing Machinery, 2023: 1007-1014.
[11] Hou, Y P, Zhang J J, Lin Z H, et al. Large Language Models are Zero-Shot Rankers for Recommender Systems[C]//Advances in Information Retrieval. Cham: Springer Nature Switzerland, 2024: 364-381.
[12] Zheng B W, Hou Y P, Lu H Y, et al. Adapting Large Language Models by Integrating Collaborative Semantics for Recommendation[C]//2024 IEEE 40th International Conference on Data Engineering (ICDE). Piscataway: IEEE, 2024: 1435-1448.
[13] 顾滢双, 桂韬, 张奇. 基于语义熵反馈强化学习的大语言模型事实性幻觉缓解[J/OL].计算机工程, 1-17[2025-07-31].https://doi.org/10.19678/j.issn.1000-3428.0252253. Gu Y S, Gui T, Zhang Q. Mitigating Factuality Hallucination in LLM with Semantic Entropy based Reinforcement Learning and Multi-Agent Collaboration[J/OL]. Computer Engineering:1-17[2025-07-31].https://doi.org/10.19678/j.issn.1000-3428.0252253. (in Chinese)
[14] Rajput S, Mehta N, Singh A, et al. Recommender Systems with Generative Retrieval[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc, 2023: 10299-10315.
[15] Liu Z H, Hou Y P, McAuley J, et al. Multi-Behavior Generative Recommendation[C]//Proceedings of the 33rd ACM International Conference on Information and Knowledge Management. New York, USA: Association for Computing Machinery, 2024: 1575-1585.
[16] Wang, Y J, Hou Y Y, Wang H N, et al. A Neural Corpus Indexer for Document Retrieval[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc, 2022: 25600-25614.
[17] Rombach R, Blattmann A, Lorenz D, et al. High-Resolution Image Synthesis With Latent Diffusion Models[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Piscataway: IEEE, 2022: 10684-10695.
[18] 李丹, 谢语涵, 韩潇帅, 等. 基于情感增强BERT与多任务生成对抗网络的虚假评论检测模型[J/OL].计算机工程,1-19[2025-07-31].https://doi.org/10.19678/j.issn.1000-3428.0252154. Li D, Xie Y H, Han X S, et al.A False Review Detection Model Based on Sentiment-Enhanced BERT and Multi-Task Generative Adversarial Networks[J/OL]. Computer Engineering, 1-19[2025-07-31]. https://doi.org/10.19678/j.issn.1000-3428.0252154. (in Chinese)
[19] Jin B B, Lian D F, Liu Z, et al. Sampling-Decomposable Generative Adversarial Recommender[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc, 2020: 22629-22639.
[20] Kingma Durk P, Salimans Tim, Jozefowicz Rafal, et al. Improved Variational Inference with Inverse Autoregressive Flow[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc, 2016: 29.
[21] Wang W J, Xu, Y Y, Feng F L , et al. Diffusion Recommender Model[C]//Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: Association for Computing Machinery, 2023: 832–841.
[22] Ho Jonathan, Jain Ajay, Abbeel Pieter, et al. Denoising Diffusion Probabilistic Models[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc, 2020: 6840-6851.
[23] Yang S, Jascha S, Diederik P. K, et al. Score-Based Generative Modeling through Stochastic Differential E-quations[DB/OL]. arXiv, 2021[2025-07-30]. https://arxiv.org/abs/2011.13456.
[24] Yang L, Zhang Z L, Song Y, et al. Diffusion Models: A Comprehensive Survey of Methods and Applications[J]. ACM Comput. Surv, 2023, 56(4): 39.
[25] Liu Q D, Yan F, Zhao X Y, et al. Diffusion Au-gmentation for Sequential Recommendation[C]//Proceedings of the 32nd ACM International Conference on I- nformation and Knowledge Management. New York, USA: Association for Computing Machinery, 2023: 1576-1586.
[26] Walker J, Zhong T, Zhang F L, et al. Recommendation via Collaborative Diffusion Generative Model[C]//Knowledge Science, Engineering and Management. Cham:Springer I-nternational Publishing, 2022: 593-605.
[27] Yang Z Y, Wu J C, Wang Z C, et al. Generate What You Prefer: Reshaping Sequential Recommendation via Guided Diffusion[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc, 2023: 24247-24261.
[28] Rafailidis, Dimitrios and Crestani Fabio. Recommenda-tion with Social Relationships via Deep Learning[C]// Proceedings of the ACM SIGIR International Confere-nce on Theory of Information Retrieval. New York, USA: Association for Computing Machinery, 2017: 151-158.
[29] Hou Y P, Zhang A, Sheng L H, et al. Generative Re-commendation Models: Progress and Directions[C]//Co-mpanion Proceedings of the ACM on Web Conference 2025. New York, USA: Association for Computing Machinery, 2025:13-16.
[30] Li S, Kawale J, Fu Y, et al. Deep Collaborative Filtering via Marginalized Denoising Auto-encoder[C]//Proceedings of the 24th ACM International on Conference on Information and Knowledge Management. New York, USA: Association for Computing Machinery, 2015: 811–820.
[31] Li X P, She J. Collaborative Variational Autoencoder for Recommender Systems[C]//Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA: Association for Computing Machinery, 2017: 305–314.
[32] Ma J X, Zhou C, Cui P, et al. Learning Disentangled Representations for Recommendation[C]//Advances in Neural Information Processing Systems. New York, USA: Curran Associates, Inc., 2019: 5712-5723.
[33] Zhang S, Yao L N, Xu X W. AutoSVD++: An Efficient Hybrid Collaborative Filtering Model via Contractive Auto-encoders[C]//Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: Association for Computing Machinery, 2017: 957–960.
[34] Li W C, Huang R, Zhao H J, et al. DimeRec: A Unified Framework for Enhanced Sequential Recommendation via Generative Diffusion Models[C]//Proceedings of the Eighteenth ACM International Conference on Web Search and Data Mining. New York, USA: Association for Computing Machinery, 2025: 726–734.
[35] Guo G B, Zhou H, Chen B W, et al. IPGAN: Generating Informative Item Pairs by Adversarial Sampling[J].IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(2): 694-706.
[36] Rendle S, Freudenthaler C, Gantner Z,et al. BPR: Bayesian Personalized Ranking from Implicit Feedback[J].AUAI Press, 2012, 452-461.
[37] He X, Deng K, Wang X,et al. LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, ACM, 2020: 639-648.
[38] Wu Y, DuBois C, Zheng A. X. Collaborative Denoising Auto-Encoders for Top-N Recommender Systems[C]//Proceedings of the Ninth ACM International Conference on Web Search and Data Mining. New York, USA: Association for Computing Machinery, 2016: 153–162.
[39] Walker J, Zhong T, Zhang F L, et al. Recommendation via Collaborative Diffusion Generative Model[C]//Kn-owledge Science, Engineering and Management. Cham:Springer International Publishing, 2022: 593-605.
[39] Ren X B, Wei W, Xia L h, et al. Representation learning with large language models for recommendation[C]//Proceedings of the ACM on Web Conference 2024. New York: ACM, 2024: 3464--3475.
[41] Li W C, Huang R, Zhao H J, et al. DimeRec: A Unified Framework for Enhanced Sequential Recommendation via Generative Diffusion Models[C]//Proceedings of the Eighteenth ACM International Conference on Web Search and Data Mining. New York: ACM, 2025: 726-734.
[40] Wan M T and McAuley Julian. Item recommendation on monotonic behavior chains[C]//Proceedings of the 12th ACM Conference on Recommender Systems. New York: ACM, 2018: 86-94.
[41] Nabiha Asghar. Yelp Dataset Challenge: Review Rating Prediction[DB/OL]. arXiv, 2016[2026-01-16]. https://arxiv.org/abs/1605.05362.
[42] Harper F. Maxwell and Konstan Joseph A. The MovieLens Datasets: History and Context[J]. ACM Transactions on Interactive Intelligent Systems, 2015, 5(4): 19.

选择文件类型/文献管理软件名称

选择包含的内容