Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Data Augmentation for Sequential Recommendation: A Survey

Yizhou Dang, Enneng Yang, Yuting Liu, Guibing Guo, Linying Jiang, Jianzhe Zhao, Xingwei Wang

TL;DR

Data sparsity challenges in sequential recommendation motivate a comprehensive survey of data augmentation (DA) methods. The paper classifies DA into heuristic-based and model-based approaches, detailing data-level and representation-level techniques, and contrasting sequence extension, refining, generation, and LLM-based strategies. Empirical analysis across multiple datasets suggests model-based augmentation often outperforms heuristic methods, though trade-offs in cost and scalability remain; hybrid and scenario-specific strategies can yield strong gains. The work highlights gaps in theory, evaluation metrics for augmented data quality, and automated, generalizable DA methods, pointing to LLM-driven augmentation as a promising frontier.

Abstract

As an essential branch of recommender systems, sequential recommendation (SR) has received much attention due to its well-consistency with real-world situations. However, the widespread data sparsity issue limits the SR model's performance. Therefore, researchers have proposed many data augmentation (DA) methods to mitigate this phenomenon and have achieved impressive progress. In this survey, we provide a comprehensive review of DA methods for SR. We start by introducing the research background and motivation. Then, we categorize existing methodologies regarding their augmentation principles, objects, and purposes. Next, we present a comparative discussion of their advantages and disadvantages, followed by the exhibition and analysis of representative experimental results. Finally, we outline directions for future research and summarize this survey. We also maintain a repository with a paper list at \url{https://github.com/KingGugu/DA-CL-4Rec}.

Data Augmentation for Sequential Recommendation: A Survey

TL;DR

Data sparsity challenges in sequential recommendation motivate a comprehensive survey of data augmentation (DA) methods. The paper classifies DA into heuristic-based and model-based approaches, detailing data-level and representation-level techniques, and contrasting sequence extension, refining, generation, and LLM-based strategies. Empirical analysis across multiple datasets suggests model-based augmentation often outperforms heuristic methods, though trade-offs in cost and scalability remain; hybrid and scenario-specific strategies can yield strong gains. The work highlights gaps in theory, evaluation metrics for augmented data quality, and automated, generalizable DA methods, pointing to LLM-driven augmentation as a promising frontier.

Abstract

As an essential branch of recommender systems, sequential recommendation (SR) has received much attention due to its well-consistency with real-world situations. However, the widespread data sparsity issue limits the SR model's performance. Therefore, researchers have proposed many data augmentation (DA) methods to mitigate this phenomenon and have achieved impressive progress. In this survey, we provide a comprehensive review of DA methods for SR. We start by introducing the research background and motivation. Then, we categorize existing methodologies regarding their augmentation principles, objects, and purposes. Next, we present a comparative discussion of their advantages and disadvantages, followed by the exhibition and analysis of representative experimental results. Finally, we outline directions for future research and summarize this survey. We also maintain a repository with a paper list at \url{https://github.com/KingGugu/DA-CL-4Rec}.
Paper Structure (18 sections, 12 equations, 4 figures, 2 tables)

This paper contains 18 sections, 12 equations, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Preliminaries
  3. Formulation of Sequential Recommendation
  4. What is Data Augmentation
  5. How Can We Use DA in SR
  6. Exploration on Our Taxonomy
  7. Heuristic-based Augmentation
  8. Basic Data-level Operators
  9. Improved Data-level Operators
  10. Representation-level Operators
  11. Model-based Augmentation
  12. Sequence Extension and Refining
  13. Sequence Generation
  14. LLM-based Augmentation
  15. Discussion
  16. ...and 3 more sections

Figures (4)

  • Figure 1: The taxonomy of data augmentation for sequential recommendation.
  • Figure 2: An illustration different training paradigms.
  • Figure 3: Basic data-level augmentation operations.
  • Figure 4: An example of sequence extension and refining methods.