[论文] Bradley-Terry Rankings for Recommender Systems Across Dataset Taxonomi...

论文概要

研究领域: ML 作者: Ekaterina Grishina, Stepan Kuznetsov, Askar Tsyganov 发布时间: 2025-06-11 arXiv: 2506.08633

中文摘要

推荐算法的排序是一个具有挑战性的问题，因为模型性能对数据集的稀疏性、序列结构和规模等特征敏感。这催生了对算法公平比较方法论的需求。简单聚合性能指标（如在基准上平均NDCG）可能产生误导性排序，破坏实际选择。为解决该问题，本文提出一种基于Bradley-Terry(BT)模型的新型数据驱动排序方法。我们证明所得排序依赖于关键数据集统计量。此外，我们提出了一种评估排序一致性的新指标，并展示了我们的排序对不完整数据的鲁棒性。最后，我们引入了一种数据集特定方法，无需运行模型即可对未见数据集上的算法进行排序，依赖于BT框架的扩展，包括BT树和带协变量的BT模型。

原文摘要

The ranking of recommendation algorithms is a challenging problem since model performance is sensitive to dataset characteristics such as sparsity, sequential structure, and scale. This drives a demand for a proper methodology for fair comparison between algorithms. Naive aggregation of performance metrics (e.g., averaging NDCG over benchmarks) can yield misleading rankings, undermining practical selection. To address this problem, we introduce a novel, data-driven ranking methodology based on Bradley-Terry (BT) model. We demonstrate that the obtained ranking depends on key dataset statistics. Additionally, we propose a novel metric for evaluating ranking consistency and demonstrate robustness of our ranking to incomplete data. Finally, we introduce a dataset-specific methodology for ranki...

--- *自动采集于 2026-06-09*

#论文 #arXiv #ML #小凯