LiRank: Industrial Large Scale Ranking Models at LinkedIn
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914904101879808 |
|---|---|
| author | Borisyuk, Fedor Zhou, Mingzhou Song, Qingquan Zhu, Siyu Tiwana, Birjodh Parameswaran, Ganesh Dangi, Siddharth Hertel, Lars Xiao, Qiang Hou, Xiaochen Ouyang, Yunbo Gupta, Aman Singh, Sheallika Liu, Dan Cheng, Hailing Le, Lei Hung, Jonathan Keerthi, Sathiya Wang, Ruoyan Zhang, Fengyu Kothari, Mohit Zhu, Chen Sun, Daqi Dai, Yun Luan, Xun Zhu, Sirou Wang, Zhiwei Daftary, Neil Shen, Qianqi Jiang, Chengming Wei, Haichao Varshney, Maneesh Ghoting, Amol Ghosh, Souvik |
| author_facet | Borisyuk, Fedor Zhou, Mingzhou Song, Qingquan Zhu, Siyu Tiwana, Birjodh Parameswaran, Ganesh Dangi, Siddharth Hertel, Lars Xiao, Qiang Hou, Xiaochen Ouyang, Yunbo Gupta, Aman Singh, Sheallika Liu, Dan Cheng, Hailing Le, Lei Hung, Jonathan Keerthi, Sathiya Wang, Ruoyan Zhang, Fengyu Kothari, Mohit Zhu, Chen Sun, Daqi Dai, Yun Luan, Xun Zhu, Sirou Wang, Zhiwei Daftary, Neil Shen, Qianqi Jiang, Chengming Wei, Haichao Varshney, Maneesh Ghoting, Amol Ghosh, Souvik |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_06859 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LiRank: Industrial Large Scale Ranking Models at LinkedIn Borisyuk, Fedor Zhou, Mingzhou Song, Qingquan Zhu, Siyu Tiwana, Birjodh Parameswaran, Ganesh Dangi, Siddharth Hertel, Lars Xiao, Qiang Hou, Xiaochen Ouyang, Yunbo Gupta, Aman Singh, Sheallika Liu, Dan Cheng, Hailing Le, Lei Hung, Jonathan Keerthi, Sathiya Wang, Ruoyan Zhang, Fengyu Kothari, Mohit Zhu, Chen Sun, Daqi Dai, Yun Luan, Xun Zhu, Sirou Wang, Zhiwei Daftary, Neil Shen, Qianqi Jiang, Chengming Wei, Haichao Varshney, Maneesh Ghoting, Amol Ghosh, Souvik Machine Learning Artificial Intelligence Information Retrieval H.3.3 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. |
| title | LiRank: Industrial Large Scale Ranking Models at LinkedIn |
| topic | Machine Learning Artificial Intelligence Information Retrieval H.3.3 |
| url | https://arxiv.org/abs/2402.06859 |