Scaling User Modeling: Large-scale Online User Representations for Ads Personalization in Meta
This work addresses scalability and efficiency problems for large-scale personalized advertising systems at companies like Meta, though it appears incremental as it builds on existing user modeling and embedding techniques.
The paper tackles the challenge of creating effective user representations for personalized advertising at Meta's scale, where constraints on training throughput, serving latency, and memory limit model complexity. It presents the Scaling User Modeling (SUM) framework, which uses upstream models to generate user embeddings shared across hundreds of ads ranking models, resulting in significant online metric gains and improved infrastructure efficiency while processing hundreds of billions of daily user requests.
Effective user representations are pivotal in personalized advertising. However, stringent constraints on training throughput, serving latency, and memory, often limit the complexity and input feature set of online ads ranking models. This challenge is magnified in extensive systems like Meta's, which encompass hundreds of models with diverse specifications, rendering the tailoring of user representation learning for each model impractical. To address these challenges, we present Scaling User Modeling (SUM), a framework widely deployed in Meta's ads ranking system, designed to facilitate efficient and scalable sharing of online user representation across hundreds of ads models. SUM leverages a few designated upstream user models to synthesize user embeddings from massive amounts of user features with advanced modeling techniques. These embeddings then serve as inputs to downstream online ads ranking models, promoting efficient representation sharing. To adapt to the dynamic nature of user features and ensure embedding freshness, we designed SUM Online Asynchronous Platform (SOAP), a latency free online serving system complemented with model freshness and embedding stabilization, which enables frequent user model updates and online inference of user embeddings upon each user request. We share our hands-on deployment experiences for the SUM framework and validate its superiority through comprehensive experiments. To date, SUM has been launched to hundreds of ads ranking models in Meta, processing hundreds of billions of user requests daily, yielding significant online metric gains and improved infrastructure efficiency.