M-scan: A Multi-Scenario Causal-driven Adaptive Network for Personalized Recommendation
M-scan is a novel multi-scenario recommendation model that explicitly extracts user interests from different scenarios and eliminates biases introduced by the scenarios themselves to enhance prediction accuracy.
Adaptive Fair Representation Learning for Personalized Fairness in Recommendations via Information Alignment
The core message of this paper is to propose a novel Adaptive Fair Representation Learning (AFRL) model that achieves personalized fairness in recommendations by treating fairness requirements as inputs and learning attribute-specific embeddings and a debiased collaborative embedding, without compromising recommendation accuracy.
NFARec: A Recommender Model that Leverages Negative Feedback for Improved Recommendations
The core message of this article is that by effectively leveraging negative feedback from user interactions, the proposed NFARec model can outperform state-of-the-art recommender systems in making high-quality recommendations.
CaDRec: A Contextualized and Debiased Recommender Model for Enhancing Recommendation Accuracy
This paper proposes CaDRec, a contextualized and debiased recommender model that effectively mitigates the over-smoothing issue in graph convolution networks (GCNs) and tackles the skewed distribution of user-item interactions caused by popularity and user-individual biases.
DRE: Generating Recommendation Explanations by Aligning Large Language Models at Data-level
Leveraging large language models' reasoning capabilities, DRE proposes a data-level alignment method to generate accurate and user-centric explanations for black-box recommendation models without modifying the recommendation system.
SARDINE: A Flexible Simulator for Studying Interactive and Dynamic Recommendation Environments
SARDINE is a flexible and interpretable recommendation simulator that can help accelerate research in interactive and data-driven recommender systems by enabling the study of various sources of complexity found in real-world recommendation environments.
Mitigating Filter Bubbles in Recommendation Systems through Yin-Yang Neutralization
The proposed Agent-based Information Neutrality (AbIN) model leverages the Yin-Yang theory to balance information perception and mitigate the filter bubble effect in recommendation systems without modifying the core recommendation algorithms.
Accelerating Matrix Factorization Training for Faster Recommendation Systems
The authors propose algorithmic methods to accelerate matrix factorization (MF) training for recommendation systems, without requiring additional computational resources. They observe fine-grained structured sparsity in the decomposed feature matrices and leverage this to dynamically prune insignificant latent factors during matrix multiplication and latent factor update, leading to significant speedups.
A Diffusion Graph Transformer Model for Enhancing Top-k Recommendation Performance
The core message of this paper is to propose a novel Diffusion Graph Transformer (DiffGT) model that effectively denoises implicit user-item interactions in recommender systems by incorporating a directional diffusion process and a graph transformer architecture.
Enhancing Personalized Recommendations through Large Language Model-Powered Text Augmentation
Leveraging the comprehensive knowledge and reasoning capabilities of large language models (LLMs), the LLM-REC framework employs diverse prompting strategies to enrich item descriptions, leading to significant improvements in personalized recommendation performance.