Sequential Data Augmentation for Generative Recommendation

2025/09/17 by Geon Lee, Lee, Geon, Bhuvesh Kumar +11 · 1 citation
Computer Science · #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Information Retrieval (cs.IR) #Machine Learning (cs.LG) #Mobile Crowdsensing and Crowdsourcing #Recommender Systems and Techniques

paper · pdf · doi:10.48550/arxiv.2509.13648

openalex publication_date 2025/09/17 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Generative recommendation plays a crucial role in personalized systems, predicting users' future interactions from their historical behavior sequences. A critical yet underexplored factor in training these models is data augmentation, the process of constructing training data from user interaction histories. By shaping the training distribution, data augmentation directly and often substantially affects model generalization and performance. Nevertheless, in much of the existing work, this process is simplified, applied inconsistently, or treated as a minor design choice, without a systematic and principled understanding of its effects. Motivated by our empirical finding that different augmentation strategies can yield large performance disparities, we conduct an in-depth analysis of how they reshape training distributions and influence alignment with future targets and generalization to unseen inputs. To systematize this design space, we propose GenPAS, a generalized and principled framework that models augmentation as a stochastic sampling process over input-target pairs with three bias-controlled steps: sequence sampling, target sampling, and input sampling. This formulation unifies widely used strategies as special cases and enables flexible control of the resulting training distribution. Our extensive experiments on benchmark and industrial datasets demonstrate that GenPAS yields superior accuracy, data efficiency, and parameter efficiency compared to existing strategies, providing practical guidance for principled training data construction in generative recommendation. Our code is available at https://github.com/snap-research/GenPAS.

Citations

• GRAM: Generative Recommendation via Semantic-aware Multi-granular Late Fusion
• Data Augmentation as Free Lunch: Exploring the Test-Time Augmentation for Sequential Recommendation
• GENIUS: A Generative Framework for Universal Multimodal Search
• OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment
• Enhancing Item Tokenization for Generative Recommendation through Self-Improvement
• Is Contrastive Learning Necessary? A Study of Data Augmentation vs Contrastive Learning in Sequential Recommendation
• Language Models As Semantic Indexers
• Qwen Technical Report
• Better Generalization with Semantic IDs: A Case Study in Ranking for Recommendations
• How Can Recommender Systems Benefit from Large Language Models: A Survey
• How Can Recommender Systems Benefit from Large Language Models: A Survey
• Recommender Systems with Generative Retrieval
• Scaling Instruction-Finetuned Language Models
• RecBole 2.0: Towards a More Up-to-Date Recommendation Library
• Data Augmentation Strategies for Improving Sequential Recommender Systems
• Contrastive Self-supervised Sequential Recommendation with Robust Augmentation
• Sequential Recommendation with Graph Neural Networks
• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Language Models are Few-Shot Learners
• Optuna: A Next-generation Hyperparameter Optimization Framework
• BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer
• Self-Attentive Sequential Recommendation
• Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding
• Attention Is All You Need
• Fusing Similarity Models with Markov Chains for Sparse Sequential Recommendation
• Session-based Recommendations with Recurrent Neural Networks
• Image-based Recommendations on Styles and Substitutes

Cited by

• DiffGRM: Diffusion-based Generative Recommendation Model

Related

• Pre-training Generative Recommender with Multi-Identifier Item Tokenization
• Why Generate When You Can Transform? Unleashing Generative Attention for Dynamic Recommendation
• Massive Memorization with Hundreds of Trillions of Parameters for Sequential Transducer Generative Recommenders
• Don't Waste It: Guiding Generative Recommenders with Structured Human Priors via Multi-head Decoding
• LoopMemGR: From Behavior Logs to Evolving Memory for Generative Recommendation