Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MCLMR: A Model-Agnostic Causal Learning Framework for Multi-Behavior Recommendation

Ranxu Zhang, Junjie Meng, Ying Sun, Ziqi Xu, Bing Yin, Hao Li, Yanyong Zhang, Chao Wang

Abstract

Multi-Behavior Recommendation (MBR) leverages multiple user interaction types (e.g., views, clicks, purchases) to enrich preference modeling and alleviate data sparsity issues in traditional single-behavior approaches. However, existing MBR methods face fundamental challenges: they lack principled frameworks to model complex confounding effects from user behavioral habits and item multi-behavior distributions, struggle with effective aggregation of heterogeneous auxiliary behaviors, and fail to align behavioral representations across semantic gaps while accounting for bias distortions. To address these limitations, we propose MCLMR, a novel model-agnostic causal learning framework that can be seamlessly integrated into various MBR architectures. MCLMR first constructs a causal graph to model confounding effects and performs interventions for unbiased preference estimation. Under this causal framework, it employs an Adaptive Aggregation module based on Mixture-of-Experts to dynamically fuse auxiliary behavior information and a Bias-aware Contrastive Learning module to align cross-behavior representations in a bias-aware manner. Extensive experiments on three real-world datasets demonstrate that MCLMR achieves significant performance improvements across various baseline models, validating its effectiveness and generality. All data and code will be made publicly available. For anonymous review, our code is available at the following the link: https://github.com/gitrxh/MCLMR.

MCLMR: A Model-Agnostic Causal Learning Framework for Multi-Behavior Recommendation

Abstract

Multi-Behavior Recommendation (MBR) leverages multiple user interaction types (e.g., views, clicks, purchases) to enrich preference modeling and alleviate data sparsity issues in traditional single-behavior approaches. However, existing MBR methods face fundamental challenges: they lack principled frameworks to model complex confounding effects from user behavioral habits and item multi-behavior distributions, struggle with effective aggregation of heterogeneous auxiliary behaviors, and fail to align behavioral representations across semantic gaps while accounting for bias distortions. To address these limitations, we propose MCLMR, a novel model-agnostic causal learning framework that can be seamlessly integrated into various MBR architectures. MCLMR first constructs a causal graph to model confounding effects and performs interventions for unbiased preference estimation. Under this causal framework, it employs an Adaptive Aggregation module based on Mixture-of-Experts to dynamically fuse auxiliary behavior information and a Bias-aware Contrastive Learning module to align cross-behavior representations in a bias-aware manner. Extensive experiments on three real-world datasets demonstrate that MCLMR achieves significant performance improvements across various baseline models, validating its effectiveness and generality. All data and code will be made publicly available. For anonymous review, our code is available at the following the link: https://github.com/gitrxh/MCLMR.

Paper Structure

This paper contains 64 sections, 16 equations, 10 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Multi-behavior Recommendation
  4. Debiased Recommendation
  5. Preliminaries
  6. Basic Notations
  7. Multi-Behavior Interaction Data
  8. Model-Agnostic Setting and Embeddings
  9. Problem Formulation
  10. Methodology
  11. Causal Preference Estimation
  12. Causal View of Multi-behavior Bias
  13. Parameterizing Model Components and Biases
  14. Bias Proxies via Generalized Behavioral Conversion Rate
  15. Parameterizing Probability Distributions
  16. ...and 49 more sections

Figures (10)

  • Figure 1: The overall framework of the MCLMR model
  • Figure 2: The proposed causal graph to illustrate the relationships among behavioral bias, true user preference, and observed interactions.
  • Figure 3: Impact of the debiasing coefficient $\gamma$. Performance is shown for HR@10 (left) and NDCG@10 (right).
  • Figure 4: Impact of the MoE expert dimension. Performance is shown for HR@10 (left) and NDCG@10 (right).
  • Figure 5: Performance of CRGCN+MCLMR and CRGCN+DCCL in active and less-active user groups. (a) the absolute performance; and (b) the relative improvements of MCLMR over DCCL. ”AU” and ”LAU” are short for the active user group and the less-active user group, respectively. In (a), bars with slash and without slash corresponds to DCCL and MCLMR, respectively. Better viewed in color.
  • ...and 5 more figures