提出可通用的推荐系统事后解释方法，分析用户/物品对推荐的影响
Model-agnostic post-hoc explainability for recommender systems
- 通过删除特定用户或物品后对比模型性能，量化其影响
- 在MovieLens和Amazon Reviews数据集上验证有效，适用于深度与传统模型
- 无需修改模型即可解释推荐结果，适合需要透明性的场景
推荐系统常依赖复杂的特征嵌入和深度学习算法，虽提升推荐质量与用户体验，但降低了系统的可解释性。本文系统地应用、调整并评估了删除诊断法在推荐场景中的表现。该方法通过比较包含与不包含特定用户或物品的模型性能，量化该观测值对推荐结果的正向或负向影响。为证明方法的通用性，研究将其应用于广泛使用的基于深度学习的神经协同过滤（NCF）和经典的协同过滤方法奇异值分解（SVD）。在MovieLens和Amazon Reviews数据集上的实验揭示了模型行为，并验证了该方法在不同推荐范式间的普适性。
原文摘要 · Abstract (English)
Recommender systems often benefit from complex feature embeddings and deep learning algorithms, which deliver sophisticated recommendations that enhance user experience, engagement, and revenue. However, these methods frequently reduce the interpretability and transparency of the system. In this research, we develop a systematic application, adaptation, and evaluation of deletion diagnostics in the recommender setting. The method compares the performance of a model to that of a similar model trained without a specific user or item, allowing us to quantify how that observation influences the recommender, either positively or negatively. To demonstrate its model-agnostic nature, the proposal is applied to both Neural Collaborative Filtering (NCF), a widely used deep learning-based recommender, and Singular Value Decomposition (SVD), a classical collaborative filtering technique. Experiments on the MovieLens and Amazon Reviews datasets provide insights into model behavior and highlight the generality of the approach across different recommendation paradigms.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。