arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.00002cs.IRcs.AIcs.LG
收藏

解耦兴趣网络用于分布外点击率预测

Disentangled Interest Network for Out-of-Distribution CTR Prediction

Tsinghua University(清华大学);Hefei University of Technology(合肥工业大学)

Yu Zheng, Chen Gao, Jianxin Chang, Yanan Niu, Yang Song, Depeng Jin, Meng Wang, Yong Li

AI总结:

DiseCTR通过解耦用户兴趣,提升分布外点击率预测的准确性和鲁棒性。

摘要:

点击率(CTR)预测,即估计用户点击给定物品的概率,是在线信息服务中的关键任务。现有方法往往做出强烈假设,即训练和测试数据来自同一分布。然而,由于用户兴趣不断变化,数据分布会随之变化,导致分布外(OOD)问题。此外,用户通常有多个兴趣,其中一些兴趣的变化速度比其他快。为此,我们提出了解耦点击率预测(DiseCTR),引入了推荐的因果视角,并解耦用户兴趣的多个方面以缓解推荐中的OOD问题。我们对CTR预测进行了因果因子分解,涉及用户兴趣、曝光模型和点击模型,基于此我们开发了深度学习实现这些三种因果机制。具体来说,我们首先设计了一个具有稀疏注意力的兴趣编码器,将原始特征映射到用户兴趣,然后引入一个弱监督的兴趣解耦器来学习独立的兴趣嵌入,这些嵌入进一步通过注意力兴趣聚合器进行整合以进行预测。在三个真实世界数据集上的实验结果表明,DiseCTR在对抗最新方法的分布外推荐中实现了最佳的准确性和鲁棒性,显著提高了AUC和GAUC超过0.02,并将logloss降低了超过13.7%。进一步分析显示,DiseCTR成功解耦了用户兴趣,这是CTR预测中分布外泛化的关键。我们已发布了代码和数据在https://github.com/DavyMorgan/DiseCTR/。

英文摘要:

Click-through rate (CTR) prediction, which estimates the probability of a user clicking on a given item, is a critical task for online information services. Existing approaches often make strong assumptions that training and test data come from the same distribution. However, the data distribution varies since user interests are constantly evolving, resulting in the out-of-distribution (OOD) issue. In addition, users tend to have multiple interests, some of which evolve faster than others. Towards this end, we propose Disentangled Click-Through Rate prediction (DiseCTR), which introduces a causal perspective of recommendation and disentangles multiple aspects of user interests to alleviate the OOD issue in recommendation. We conduct a causal factorization of CTR prediction involving user interest, exposure model, and click model, based on which we develop a deep learning implementation for these three causal mechanisms. Specifically, we first design an interest encoder with sparse attention which maps raw features to user interests, and then introduce a weakly supervised interest disentangler to learn independent interest embeddings, which are further integrated by an attentive interest aggregator for prediction. Experimental results on three real-world datasets show that DiseCTR achieves the best accuracy and robustness in OOD recommendation against state-of-the-art approaches, significantly improving AUC and GAUC by over 0.02 and reducing logloss by over 13.7%. Further analyses demonstrate that DiseCTR successfully disentangles user interests, which is the key to OOD generalization for CTR prediction. We have released the code and data at https://github.com/DavyMorgan/DiseCTR/.

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。

↑