面向排序学习的概率分布优化模型
Probability distribution optimization model for learning to rank
赵峰叙 1王健 1林原 2林鸿飞1
作者信息
- 1. 大连理工大学计算机科学与技术学院,辽宁 大连 116024
- 2. 大连理工大学公共管理学院,辽宁 大连 116024
- 折叠
摘要
现有的排序学习模型依赖于模型输出的评分来表示文档间的偏序关系.考虑到这种将评分看作单一数值的限制,提出一种概率分布排序学习模型优化方法,引入排序分数的不确定性,以概率分布的形式对排序分数进行平滑,进而将排序分数大小的比较变成对分数偏序关系的概率估计.在此基础上,将该方法应用于排序学习模型 RankNet、LambdaRank 以及LambdaMART,更合理地拟合模型概率与目标概率之间的差距,从而对排序学习模型进行优化,并在多个大规模真实数据集上进行实验.结果表明,经过优化后的模型性能相比于优化前具有显著提高,验证本文方法的有效性.
Abstract
Existing learning to rank(LTR)models rely on the scores output by models to represent the partial order among docu-ments.Considering the limitation of treating scores as deterministic values,this paper proposes a probability distribution optimization method for the LTR model,which introduces the uncertainty of the ranking score.It smooths the scores in the form of probability distributions,thereby transforming the comparison of ranking scores into the probability estimation of score partial orders.The pro-posed method is applied to LTR models such as RankNet,LambdaRank,and LambdaMART.It effectively bridges the gap between the modeled probability and the target probability,leading to optimization of the LTR models.The paper conducts experiments on multiple large-scale real datasets,and the experimental results show that the optimized models outperform the original ones,which validates the effectiveness of the proposed method.
关键词
信息检索/排序学习/概率分布Key words
information retrieval/learning to rank/probability distribution引用本文复制引用
基金项目
国家自然科学基金资助项目(61976036)
出版年
2024
NETL
NSTL
万方数据