COPR: Consistency-Oriented Pre-Ranking for Online Advertising

Cascading architecture has been widely adopted in large-scale advertising systems to balance efficiency and effectiveness. In this architecture, the pre-ranking model is expected to be a lightweight approximation of the ranking model, which handles more candidates with strict latency requirements. Due to the gap in model capacity, the pre-ranking and ranking models usually generate inconsistent ranked results, thus hurting the overall system effectiveness. The paradigm of score alignment is proposed to regularize their raw scores to be consistent. However, it suffers from inevitable alignment errors and error amplification by bids when applied in online advertising. To this end, we introduce a consistency-oriented pre-ranking framework for online advertising, which employs a chunk-based sampling module and a plug-and-play rank alignment module to explicitly optimize consistency of ECPM-ranked results. A ΔNDCG-based weighting mechanism is adopted to better distinguish the importance of inter-chunk samples in optimization. Both online and offline experiments have validated the superiority of our framework. When deployed in Taobao display advertising system, it achieves an improvement of up to +12.3% CTR and +5.6% RPM.

• Publication year

  2023

• Venue

  International Conference on Information and Knowledge Management

• Publication date

  2023-06-06

• Fields of study

  Business, Computer Science

• Identifiers
  DOI 10.1145/3583780.3615465arXiv 2306.03516
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Rec4Ad: A Free Lunch to Mitigate Sample Selection Bias for Ads CTR Prediction in Taobao

  2023cited by this paper
• Capturing Conversion Rate Fluctuation during Sales Promotions: A Novel Historical Data Reuse Approach

  2023cited by this paper
• PICASSO: Unleashing the Potential of GPU-centric Training for Wide-and-deep Recommender Systems

  2022cited by this paper
• RankFlow: Joint Optimization of Multi-Stage Cascade Ranking Systems as Flows

  2022cited by this paper
• AutoFAS: Automatic Feature and Architecture Selection for Pre-Ranking System

  2022cited by this paper
• IntTower: The Next Generation of Two-Tower Model for Pre-Ranking System

  2022cited by this paper
• KEEP: An Industrial Pre-Training Framework for Online Recommendation via Knowledge Extraction and Plugging

  2022cited by this paper
• Joint Optimization of Ranking and Calibration with Contextualized Hybrid Model

  2022cited by this paper
• An Online Multi-task Learning Framework for Google Feed Ads Auction Models

  2022cited by this paper
• On Designing a Two-stage Auction for Online Advertising

  2021cited by this paper
• Towards a Better Tradeoff between Effectiveness and Efficiency in Pre-Ranking: A Learnable Feature Selection based Approach

  2021cited by this paper
• An Embedding Learning Framework for Numerical Features in CTR Prediction

  2020cited by this paper
• AutoFIS: Automatic Feature Interaction Selection in Factorization Models for Click-Through Rate Prediction

  2020cited by this paper
• Search-based User Interest Modeling with Lifelong Sequential Behavior Data for Click-Through Rate Prediction

  2020cited by this paper
• COLD: Towards the Next Generation of Pre-Ranking System

  2020influential reference
• Improving Ad Click Prediction by Considering Non-displayed Events

  2019cited by this paper
• TF-Ranking: Scalable TensorFlow Library for Learning-to-Rank

  2018cited by this paper
• Ranking Distillation: Learning Compact Ranking Models With High Performance for Recommender System

  2018cited by this paper
• EENMF: An End-to-End Neural Matching Framework for E-Commerce Sponsored Search

  2018cited by this paper
• Deep Interest Network for Click-Through Rate Prediction

  2017cited by this paper
• Squeeze-and-Excitation Networks

  2017cited by this paper
• Deep Neural Networks for YouTube Recommendations

  2016cited by this paper
• Real-Time Bidding: A New Frontier of Computational Advertising Research

  2015cited by this paper
• Distilling the Knowledge in a Neural Network

  2015cited by this paper
• Learning deep structured semantic models for web search using clickthrough data

  2013cited by this paper
• Online Display Advertising: Targeting and Obtrusiveness

  2011cited by this paper
• From RankNet to LambdaRank to LambdaMART: An Overview

  2010cited by this paper
• Learning to rank for information retrieval

  2009cited by this paper
• Learning to rank using gradient descent

  2005cited by this paper
• IR evaluation methods for retrieving highly relevant documents

  2000cited by this paper
• IR evaluation methods for retrieving highly relevant documents

  2000influential reference

• Deep Learning to Rank in Industrial Search Engines, Recommender Systems and Online Advertising: An Overview and New Perspectives

  2026influential citation
• Not All Candidates are Created Equal: A Heterogeneity-Aware Approach to Pre-ranking in Recommender Systems

  2026cites this paper
• A Market-Aware Real-Time Bidding Strategy Using Censored Data With Reinforcement Learning in Online Advertising

  2025cites this paper
• EGA-V2: An End-to-end Generative Framework for Industrial Advertising

  2025cites this paper
• EGA-V1: Unifying Online Advertising with End-to-End Learning

  2025cites this paper
• Bidding-Aware Retrieval for Multi-Stage Consistency in Online Advertising

  2025influential citation
• Denoising Neural Reranker for Recommender Systems

  2025cites this paper
• MTMD: A Multi-Task Multi-Domain Framework for Unified Ad Lightweight Ranking at Pinterest

  2025cites this paper
• UniROM: Unifying Online Advertising Ranking as One Model

  2025cites this paper
• AIF: Asynchronous Inference Framework for Cost-Effective Pre-Ranking

  2025cites this paper
• MaRCA: Multi-Agent Reinforcement Learning for Dynamic Computation Allocation in Large-Scale Recommender Systems

  2025cites this paper
• A Hybrid Cross-Stage Coordination Pre-ranking Model for Online Recommendation Systems

  2025cites this paper
• An Unbiased Entire-Space Causal Framework for Click-Through Rate Estimation in Pre-Ranking

  2025cites this paper
• Computational Advertising: Recent Advances

  2025cites this paper
• Optimizing E-commerce Search: Toward a Generalizable and Rank-Consistent Pre-Ranking Model

  2024cites this paper
• RankTower: A Synergistic Framework for Enhancing Two-Tower Pre-Ranking Model

  2024cites this paper
• Enhancing Taobao Display Advertising with Multimodal Representations: Challenges, Approaches and Insights

  2024cites this paper
• Truthful Bandit Mechanisms for Repeated Two-stage Ad Auctions

  2024cites this paper
• A Unified Knowledge-Distillation and Semi-Supervised Learning Framework to Improve Industrial Ads Delivery Systems

  2024cites this paper
• Multi-objective Learning to Rank by Model Distillation

  2024cites this paper
• Rec4Ad: A Free Lunch to Mitigate Sample Selection Bias for Ads CTR Prediction in Taobao

  2023cites this paper
• Calibration-compatible Listwise Distillation of Privileged Features for CTR Prediction

  2023cites this paper