arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2207.02468cs.IR
收藏

模型无关匹配的负样本重加权

Re-weighting Negative Samples for Model-Agnostic Matching

Jiazhen Lou, Hong Wen, Fuyu Lv, Jing Zhang, Tengfei Yuan, Zhao Li

摘要:

推荐系统作为一种从庞大的语料库中发现用户感兴趣项目的有效工具,越来越受到学术界和工业界的关注。大规模匹配作为遥感的初始阶段,是基础性的,也是具有挑战性的。一个典型的方法是使用双塔架构学习用户和项目表示,然后计算两个表示向量之间的相似性得分,然而,这仍然在如何正确处理负样本方面存在困难。在本文中,我们发现,通常的做法,从整个空间中随机抽样的负样本,并平等对待他们不是一个最佳的选择,因为负样本从不同的子空间在不同的阶段有不同的重要性,一个匹配模型。为了解决这个问题,我们提出了一种新的方法命名为无偏模型无关匹配方法(UMA $^2$)。它由两个基本模块组成,包括1)通用匹配模型(GMM),它是模型不可知的,可以实现为任何基于嵌入的双塔模型; 2)负样本去偏网络(NSDN),它通过借用逆倾向加权(IPW)的思想来区分负样本,并重新权衡GMM中的损失。UMA $^2$将这两个模块无缝集成到一个端到端的多任务学习框架中。在真实世界的离线数据集和在线A/B测试上的大量实验证明了其优越性。

英文摘要:

Recommender Systems (RS), as an efficient tool to discover users' interested items from a very large corpus, has attracted more and more attention from academia and industry. As the initial stage of RS, large-scale matching is fundamental yet challenging. A typical recipe is to learn user and item representations with a two-tower architecture and then calculate the similarity score between both representation vectors, which however still struggles in how to properly deal with negative samples. In this paper, we find that the common practice that randomly sampling negative samples from the entire space and treating them equally is not an optimal choice, since the negative samples from different sub-spaces at different stages have different importance to a matching model. To address this issue, we propose a novel method named Unbiased Model-Agnostic Matching Approach (UMA$^2$). It consists of two basic modules including 1) General Matching Model (GMM), which is model-agnostic and can be implemented as any embedding-based two-tower models; and 2) Negative Samples Debias Network (NSDN), which discriminates negative samples by borrowing the idea of Inverse Propensity Weighting (IPW) and re-weighs the loss in GMM. UMA$^2$ seamlessly integrates these two modules in an end-to-end multi-task learning framework. Extensive experiments on both real-world offline dataset and online A/B test demonstrate its superiority over state-of-the-art methods.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑