arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.13643cs.IR
收藏

重构CTR预测:学习用于推荐的不变特征交互

Reformulating CTR Prediction: Learning Invariant Feature Interactions for Recommendation

Yang Zhang, Tianhao Shi, Fuli Feng, Wenjie Wang, Dingxian Wang, Xiangnan He, Yongdong Zhang

AI总结:

针对传统ERM范式下CTR预测学习的特征交互泛化性差的问题,提出解耦不变学习DIL及高效版本LightDIL,通过解耦特征嵌入捕获环境不变与特定相关性,实验验证了其学习稳定特征交互的有效性。

摘要:

点击率(CTR)预测在推荐系统中处于核心地位,作为最终阶段的过滤器为用户对物品进行排序。解决CTR任务的关键在于学习对预测有用的特征交互,这通常通过经验风险最小化(ERM)范式拟合历史点击数据来实现。代表性方法包括因子分解机(Factorization Machines)和深度兴趣网络(Deep Interest Network),已在工业应用中取得广泛成功。然而,这种方式不可避免地会学习到不稳定的特征交互,即那些在历史数据中表现出强相关性但在未来服务中泛化能力很差的交互。\n在本研究中,我们重构了CTR任务——不再在历史数据上追求ERM,而是将历史数据按时间顺序划分为若干时段(也称为环境),旨在学习跨时段稳定的特征交互。这类特征交互理应具备更好的泛化能力,可用于预测未来行为数据。不过,一个技术挑战在于现有的不变学习解决方案(如不变风险最小化)并不适用,因为点击数据同时混杂了环境不变相关性和环境特定相关性。为解决这一难题,我们提出了解耦不变学习(Disentangled Invariant Learning,DIL),该方法对特征嵌入进行解耦,以分别捕获这两类相关性。为提升建模效率,我们进一步设计了LightDIL,它在特征域的更高层级执行解耦。大量实验证明了DIL在为CTR学习稳定特征交互方面的有效性。我们已在https://github.com/zyang1580/DIL 发布代码。

英文摘要:

Click-Through Rate (CTR) prediction plays a core role in recommender systems, serving as the final-stage filter to rank items for a user. The key to addressing the CTR task is learning feature interactions that are useful for prediction, which is typically achieved by fitting historical click data with the Empirical Risk Minimization (ERM) paradigm. Representative methods include Factorization Machines and Deep Interest Network, which have achieved wide success in industrial applications. However, such a manner inevitably learns unstable feature interactions, i.e., the ones that exhibit strong correlations in historical data but generalize poorly for future serving. In this work, we reformulate the CTR task -- instead of pursuing ERM on historical data, we split the historical data chronologically into several periods (a.k.a, environments), aiming to learn feature interactions that are stable across periods. Such feature interactions are supposed to generalize better to predict future behavior data. Nevertheless, a technical challenge is that existing invariant learning solutions like Invariant Risk Minimization are not applicable, since the click data entangles both environment-invariant and environment-specific correlations. To address this dilemma, we propose Disentangled Invariant Learning (DIL) which disentangles feature embeddings to capture the two types of correlations separately. To improve the modeling efficiency, we further design LightDIL which performs the disentanglement at the higher level of the feature field. Extensive experiments demonstrate the effectiveness of DIL in learning stable feature interactions for CTR. We release the code at https://github.com/zyang1580/DIL.

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。

↑