AI

ACQ: A Deployed Two-Stage Framework for Automated Creative Quota Allocation in Large-Scale Online Advertising

Researchers have developed Automated Creatives Quota (ACQ), a framework for determining the optimal number of ad creatives to generate from a single photo in online advertising. ACQ consists of two stages: predicting expected revenue based on quota levels using a multi-task model, and solving a multiple-choice knapsack problem to allocate quotas under global capacity constraints. The framework was tested on Kuaishou's advertising platform and found to increase advertising rev
Researchers have developed Automated Creatives Quota (ACQ), a framework for determining the optimal number of ad creatives to generate from a single photo in online advertising. ACQ consists of two stages: predicting expected revenue based on quota levels using a multi-task model, and solving a multiple-choice knapsack problem to allocate quotas under global capacity constraints. The framework was tested on Kuaishou's advertising platform and found to increase advertising revenue by 6.20%. The authors claim that ACQ can handle large-scale online advertising scenarios more efficiently than existing methods. --- Why it matters: This matters because it addresses a practical problem in digital advertising, where generating too many ad creatives can lead to diminishing returns on investment. Engineers working on AI-powered advertising platforms may be interested in implementing similar quota allocation frameworks to optimize revenue. Source: https://arxiv.org/abs/2412.06167

This article was originally published at: https://arxiv.org/abs/2412.06167