Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Causal Learning for Trustworthy Recommender Systems: A Survey

Jin Li, Shoujin Wang, Qi Zhang, Longbing Cao, Fang Chen, Xiuzhen Zhang, Dietmar Jannach, Charu C. Aggarwal

TL;DR

Trustworthy recommender systems require fairness, robustness, and explainability beyond accuracy. The paper advocates causality-oriented TRS (CTRS) and presents a two-step framework of causal formulation and causal inference. It provides a granular, stage-wise taxonomy linking trustworthiness challenges to causal solutions and classifies CTRS methods into causal discovery, causal effect inference, and counterfactual reasoning, with open problems and future directions. The work emphasizes causal-oriented evaluation and the integration of generative models and large language models with bias mitigation to improve real-world applicability.

Abstract

Recommender Systems (RS) have significantly advanced online content filtering and personalized decision-making. However, emerging vulnerabilities in RS have catalyzed a paradigm shift towards Trustworthy RS (TRS). Despite substantial progress on TRS, most efforts focus on data correlations while overlooking the fundamental causal nature of recommendations. This drawback hinders TRS from identifying the root cause of trustworthiness issues, leading to limited fairness, robustness, and explainability. To bridge this gap, causal learning emerges as a class of promising methods to augment TRS. These methods, grounded in reliable causality, excel in mitigating various biases and noise while offering insightful explanations for TRS. However, there is a lack of timely and dedicated surveys in this vibrant area. This paper creates an overview of TRS from the perspective of causal learning. We begin by presenting the advantages and common procedures of Causality-oriented TRS (CTRS). Then, we identify potential trustworthiness challenges at each stage and link them to viable causal solutions, followed by a classification of CTRS methods. Finally, we discuss several future directions for advancing this field.

Causal Learning for Trustworthy Recommender Systems: A Survey

TL;DR

Trustworthy recommender systems require fairness, robustness, and explainability beyond accuracy. The paper advocates causality-oriented TRS (CTRS) and presents a two-step framework of causal formulation and causal inference. It provides a granular, stage-wise taxonomy linking trustworthiness challenges to causal solutions and classifies CTRS methods into causal discovery, causal effect inference, and counterfactual reasoning, with open problems and future directions. The work emphasizes causal-oriented evaluation and the integration of generative models and large language models with bias mitigation to improve real-world applicability.

Abstract

Recommender Systems (RS) have significantly advanced online content filtering and personalized decision-making. However, emerging vulnerabilities in RS have catalyzed a paradigm shift towards Trustworthy RS (TRS). Despite substantial progress on TRS, most efforts focus on data correlations while overlooking the fundamental causal nature of recommendations. This drawback hinders TRS from identifying the root cause of trustworthiness issues, leading to limited fairness, robustness, and explainability. To bridge this gap, causal learning emerges as a class of promising methods to augment TRS. These methods, grounded in reliable causality, excel in mitigating various biases and noise while offering insightful explanations for TRS. However, there is a lack of timely and dedicated surveys in this vibrant area. This paper creates an overview of TRS from the perspective of causal learning. We begin by presenting the advantages and common procedures of Causality-oriented TRS (CTRS). Then, we identify potential trustworthiness challenges at each stage and link them to viable causal solutions, followed by a classification of CTRS methods. Finally, we discuss several future directions for advancing this field.
Paper Structure (29 sections, 6 equations, 2 figures, 2 tables)

This paper contains 29 sections, 6 equations, 2 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Motivation: why Causality-oriented TRS (CTRS)?
  3. How does causal learning help solve TRS problems?
  4. Comparison with Existing Surveys
  5. Trustworthiness Challenges in RS Lifecycle
  6. Problem Formulation
  7. Challenges in Data Preparation
  8. Fairness Challenges in Data Preparation
  9. Robustness Challenges in Data Preparation
  10. Challenges in Representation Learning
  11. Fairness Challenges in Representation Learning
  12. Robustness Challenges in Representation Learning
  13. Challenges in Recommendation Generation
  14. Fairness Challenges in Recommendation Generation
  15. Explainability Challenges in Recommendation
  16. ...and 14 more sections

Figures (2)

  • Figure 1: Conventional RS relies on correlations, often misled by popularity bias (e.g., suggesting trending thrillers to users who prefer period dramas). CTRS performs causal interventions to reduce bias, ensuring more relevant and trustworthy recommendations.
  • Figure 2: Classification of CTRS methods.