Data Augmentation as Free Lunch: Exploring the Test-Time Augmentation for Sequential Recommendation

2025-04-07

AI Summary

To address data sparsity in sequential recommendation, existing data augmentation methods rely on retraining or model modifications—rendering them incompatible with deployed large-scale models. This paper pioneers the application of test-time augmentation (TTA) to sequential recommendation: lightweight perturbations are applied to user behavior sequences during inference, and predictions from multiple augmented views are ensembled—requiring no backpropagation, retraining, or architectural changes. We propose two efficient TTA operators: TNoise, which injects uniform noise into item embeddings, and TMask, which applies computationally lightweight masking without item replacement—preserving sequential patterns while avoiding the computational overhead and semantic distortion inherent in conventional masking or item-replacement strategies. Extensive experiments across multiple benchmark datasets demonstrate significant improvements in Recall@20 and NDCG@20, with inference latency increase under 3%. The approach is model-agnostic and compatible with diverse sequential recommenders. Code is publicly available.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Data Mining & Knowledge Management: Recommender Systems
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Data augmentation has become a promising method of mitigating data sparsity in sequential recommendation. Existing methods generate new yet effective data during model training to improve performance. However, deploying them requires retraining, architecture modification, or introducing additional learnable parameters. The above steps are time-consuming and costly for well-trained models, especially when the model scale becomes large. In this work, we explore the test-time augmentation (TTA) for sequential recommendation, which augments the inputs during the model inference and then aggregates the model's predictions for augmented data to improve final accuracy. It avoids significant time and cost overhead from loss calculation and backward propagation. We first experimentally disclose the potential of existing augmentation operators for TTA and find that the Mask and Substitute consistently achieve better performance. Further analysis reveals that these two operators are effective because they retain the original sequential pattern while adding appropriate perturbations. Meanwhile, we argue that these two operators still face time-consuming item selection or interference information from mask tokens. Based on the analysis and limitations, we present TNoise and TMask. The former injects uniform noise into the original representation, avoiding the computational overhead of item selection. The latter blocks mask token from participating in model calculations or directly removes interactions that should have been replaced with mask tokens. Comprehensive experiments demonstrate the effectiveness, efficiency, and generalizability of our method. We provide an anonymous implementation at https://github.com/KingGugu/TTA4SR.

Problem

• Mitigates data sparsity in sequential recommendation
• Avoids retraining and reduces time-costly overhead
• Improves accuracy with test-time augmentation techniques

Innovation

• Test-time augmentation for sequential recommendation
• Uniform noise injection avoids item selection
• Mask token removal improves model efficiency