LiRank: Industrial Large Scale Ranking Models at LinkedIn
This work addresses the challenge of deploying efficient and effective ranking systems for industrial applications like social media feeds, job recommendations, and ads, though it is incremental as it builds on existing methods.
The authors tackled the problem of improving large-scale ranking models at LinkedIn by developing LiRank, a framework that integrates state-of-the-art architectures and optimization methods, resulting in relative metrics improvements such as +0.5% member sessions in Feed, +1.76% qualified job applications, and +4.3% Ads CTR.
We present LiRank, a large-scale ranking framework at LinkedIn that brings to production state-of-the-art modeling architectures and optimization methods. We unveil several modeling improvements, including Residual DCN, which adds attention and residual connections to the famous DCNv2 architecture. We share insights into combining and tuning SOTA architectures to create a unified model, including Dense Gating, Transformers and Residual DCN. We also propose novel techniques for calibration and describe how we productionalized deep learning based explore/exploit methods. To enable effective, production-grade serving of large ranking models, we detail how to train and compress models using quantization and vocabulary compression. We provide details about the deployment setup for large-scale use cases of Feed ranking, Jobs Recommendations, and Ads click-through rate (CTR) prediction. We summarize our learnings from various A/B tests by elucidating the most effective technical approaches. These ideas have contributed to relative metrics improvements across the board at LinkedIn: +0.5% member sessions in the Feed, +1.76% qualified job applications for Jobs search and recommendations, and +4.3% for Ads CTR. We hope this work can provide practical insights and solutions for practitioners interested in leveraging large-scale deep ranking systems.