Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-Behavior Recommender Systems: A Survey

Kyungho Kim, Sunwoo Kim, Geon Lee, Jinhong Jung, Kijung Shin

TL;DR

This survey addresses how to integrate multiple user behaviors into recommender systems by proposing a three-step framework: data modeling, encoding, and training. It provides a comprehensive taxonomy of data representations (view-specific graphs, view-unified graphs, and view-unified sequences), encoding strategies (parallel and sequential), and training objectives (main and auxiliary). The work catalogs benchmark datasets across implicit and explicit feedback settings and contrasts multiple methodological approaches, including graph-based, sequence-based, and knowledge-augmented methods. By outlining challenges and future directions—such as data sparsity, scalability, temporal dynamics, interpretability, and privacy—the paper offers a holistic guide for researchers and practitioners seeking to design more accurate and robust multi-behavior recommender systems.

Abstract

Traditional recommender systems primarily rely on a single type of user-item interaction, such as item purchases or ratings, to predict user preferences. However, in real-world scenarios, users engage in a variety of behaviors, such as clicking on items or adding them to carts, offering richer insights into their interests. Multi-behavior recommender systems leverage these diverse interactions to enhance recommendation quality, and research on this topic has grown rapidly in recent years. This survey provides a timely review of multi-behavior recommender systems, focusing on three key steps: (1) Data Modeling: representing multi-behaviors at the input level, (2) Encoding: transforming these inputs into vector representations (i.e., embeddings), and (3) Training: optimizing machine-learning models. We systematically categorize existing multi-behavior recommender systems based on the commonalities and differences in their approaches across the above steps. Additionally, we discuss promising future directions for advancing multi-behavior recommender systems.

Multi-Behavior Recommender Systems: A Survey

TL;DR

This survey addresses how to integrate multiple user behaviors into recommender systems by proposing a three-step framework: data modeling, encoding, and training. It provides a comprehensive taxonomy of data representations (view-specific graphs, view-unified graphs, and view-unified sequences), encoding strategies (parallel and sequential), and training objectives (main and auxiliary). The work catalogs benchmark datasets across implicit and explicit feedback settings and contrasts multiple methodological approaches, including graph-based, sequence-based, and knowledge-augmented methods. By outlining challenges and future directions—such as data sparsity, scalability, temporal dynamics, interpretability, and privacy—the paper offers a holistic guide for researchers and practitioners seeking to design more accurate and robust multi-behavior recommender systems.

Abstract

Traditional recommender systems primarily rely on a single type of user-item interaction, such as item purchases or ratings, to predict user preferences. However, in real-world scenarios, users engage in a variety of behaviors, such as clicking on items or adding them to carts, offering richer insights into their interests. Multi-behavior recommender systems leverage these diverse interactions to enhance recommendation quality, and research on this topic has grown rapidly in recent years. This survey provides a timely review of multi-behavior recommender systems, focusing on three key steps: (1) Data Modeling: representing multi-behaviors at the input level, (2) Encoding: transforming these inputs into vector representations (i.e., embeddings), and (3) Training: optimizing machine-learning models. We systematically categorize existing multi-behavior recommender systems based on the commonalities and differences in their approaches across the above steps. Additionally, we discuss promising future directions for advancing multi-behavior recommender systems.

Paper Structure

This paper contains 51 sections, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Step 1: Data Modeling.
  3. Step 2: Encoding.
  4. Step 3: Training.
  5. Comparison with Existing Surveys.
  6. Preliminaries for Multi-behavior Recommendation
  7. Task Description
  8. Overview.
  9. Key Steps in Multi-behavior Recommendation.
  10. Benchmark Datasets for Multi-behavior Recommendation
  11. Dataset Categorization.
  12. Dataset Summaries.
  13. Datasets without explicit feedback:
  14. Datasets with explicit feedback:
  15. Step 1. Data Modeling
  16. ...and 36 more sections

Figures (2)

  • Figure 1: Multi-behavior recommendation in two example domains.
  • Figure 2: Overview of key components in multi-behavior recommendation.