Dually Enhanced Delayed Feedback Modeling for Streaming Conversion Rate Prediction
Open Access
- 21 October 2023
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 390-399
Abstract
In online industrial advertising systems, conversion actions (e.g., purchases or downloads) often occur significantly delayed, even up to several days or weeks after the user clicks. This phenomenon leads to the crucial challenge calleddelayed feedback problem in streaming CVR prediction, that is, the online systems cannot receive the true label of conversions immediately for continuous training. To mitigate the delayed feedback problem, recent state-of-the-art methods often apply sample duplicate mechanisms to introduce early certain conversion information. Nevertheless, these works have overlooked a crucial issue of rapid shifts in data distribution and considered both the newly observed data and duplicated early data together, resulting in biases in both distributions. In this work, we propose a Dually enhanced Delayed Feedback Model (DDFM), which tackles the above issues by treating the newly observed data and duplicated early data separately. DDFM consists of dual unbiased CVR estimators that share the same form but utilize different latent variables as weights: one for the newly observed data and the other for the duplicated early data. To avoid high variance, we adopt an addition-only formula for these latent variables, eliminating multiplication or division operations. Furthermore, we design a shared-bottom network that efficiently and jointly estimates the latent variables in DDFM. Theoretical analysis demonstrates the unbiasedness and convergence properties of DDFM. Extensive experiments on both public and industrial large-scale real-world datasets exhibit that our proposed DDFM consistently outperforms existing state-of-the-art methods.Keywords
Funding Information
- fund for building world-class universities (disciplines) of Renmin University of China
- Intelligent Social Governance Interdisciplinary Platform
- Major Innovation & Planning Interdisciplinary Platform for the ``Double-First Class' Initiative, Renmin University of China
- National Key Research and Development Program of China (2019YFE0198200)
- Beijing Outstanding Young Scientist Program (BJJWZYJH012019100020098)
This publication has 28 references indexed in Scilit:
- Streaming Session-based Recommendation2019
- Streamingrec2018
- Deep Interest Network for Click-Through Rate Prediction2018
- Entire Space Multi-Task Model2018
- A Practical Framework of Conversion Rate Prediction for Online Display Advertising2017
- Optimized Cost per Click in Taobao Display Advertising2017
- Causal Inference for Statistics, Social, and Biomedical Sciences2015
- Modeling delayed feedback in display advertising2014
- Importance sampling: a reviewWIREs Computational Statistics, 2009
- Multi-armed Bandit Algorithms and Empirical Evaluation2005