Buchumschlag

Meta Lattice: Model Space Redesign for Cost-Effective Industry-Scale Ads Recommendations

Fuente: arXiv

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	Luo, Liang, Chen, Yuxin, Zhang, Zhengyu, Hang, Mengyue, Gu, Andrew, Zhang, Buyun, Liu, Boyang, Chen, Chen, Fan, Chengze, Liang, Dong, Yang, Fan, Gu, Feifan, Li, Huayu, Nie, Jade, Xu, Jiayi, Yang, Jiyan, Park, Jongsoo, Chen, Laming, Jin, Longhao, Li, Qianru, Huang, Qin, Jiang, Shali, Shen, Shiwen, Wang, Shuaiwen, Zeng, Sihan, Yuan, Siyang, Tang, Tongyi, Zhang, Weilin, Wang, Wenjun, Liu, Xi, Wei, Xiaohan, Xia, Xiaozhen, Hao, Yuchen, He, Yunlong, Badr, Yasmine, Chen, Zeliang, Naumov, Maxim, Yao, Yantao, Chen, Wenlin, Kolay, Santanu, Musumeci, GP, Wen, Ellie Dingqiao
Format:	Preprint
Veröffentlicht:	2025
Schlagworte:	Information Retrieval
Online-Zugang:	Acceder al recurso
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

_version_	1866908711176372224
author	Luo, Liang Chen, Yuxin Zhang, Zhengyu Hang, Mengyue Gu, Andrew Zhang, Buyun Liu, Boyang Chen, Chen Fan, Chengze Liang, Dong Yang, Fan Gu, Feifan Li, Huayu Nie, Jade Xu, Jiayi Yang, Jiyan Park, Jongsoo Chen, Laming Jin, Longhao Li, Qianru Huang, Qin Jiang, Shali Shen, Shiwen Wang, Shuaiwen Zeng, Sihan Yuan, Siyang Tang, Tongyi Zhang, Weilin Wang, Wenjun Liu, Xi Wei, Xiaohan Xia, Xiaozhen Hao, Yuchen He, Yunlong Badr, Yasmine Chen, Zeliang Naumov, Maxim Yao, Yantao Chen, Wenlin Kolay, Santanu Musumeci, GP Wen, Ellie Dingqiao
author_facet	Luo, Liang Chen, Yuxin Zhang, Zhengyu Hang, Mengyue Gu, Andrew Zhang, Buyun Liu, Boyang Chen, Chen Fan, Chengze Liang, Dong Yang, Fan Gu, Feifan Li, Huayu Nie, Jade Xu, Jiayi Yang, Jiyan Park, Jongsoo Chen, Laming Jin, Longhao Li, Qianru Huang, Qin Jiang, Shali Shen, Shiwen Wang, Shuaiwen Zeng, Sihan Yuan, Siyang Tang, Tongyi Zhang, Weilin Wang, Wenjun Liu, Xi Wei, Xiaohan Xia, Xiaozhen Hao, Yuchen He, Yunlong Badr, Yasmine Chen, Zeliang Naumov, Maxim Yao, Yantao Chen, Wenlin Kolay, Santanu Musumeci, GP Wen, Ellie Dingqiao
contents	The rapidly evolving landscape of products, surfaces, policies, and regulations poses significant challenges for deploying state-of-the-art recommendation models at industry scale, primarily due to data fragmentation across domains and escalating infrastructure costs that hinder sustained quality improvements. To address this challenge, we propose Lattice, a recommendation framework centered around model space redesign that extends Multi-Domain, Multi-Objective (MDMO) learning beyond models and learning objectives. Lattice addresses these challenges through a comprehensive model space redesign that combines cross-domain knowledge sharing, data consolidation, model unification, distillation, and system optimizations to achieve significant improvements in both quality and cost-efficiency. Our deployment of Lattice at Meta has resulted in 10% revenue-driving top-line metrics gain, 11.5% user satisfaction improvement, 6% boost in conversion rate, with 20% capacity saving.
format	Preprint
id	arxiv_https___arxiv_org_abs_2512_09200
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	Meta Lattice: Model Space Redesign for Cost-Effective Industry-Scale Ads Recommendations Luo, Liang Chen, Yuxin Zhang, Zhengyu Hang, Mengyue Gu, Andrew Zhang, Buyun Liu, Boyang Chen, Chen Fan, Chengze Liang, Dong Yang, Fan Gu, Feifan Li, Huayu Nie, Jade Xu, Jiayi Yang, Jiyan Park, Jongsoo Chen, Laming Jin, Longhao Li, Qianru Huang, Qin Jiang, Shali Shen, Shiwen Wang, Shuaiwen Zeng, Sihan Yuan, Siyang Tang, Tongyi Zhang, Weilin Wang, Wenjun Liu, Xi Wei, Xiaohan Xia, Xiaozhen Hao, Yuchen He, Yunlong Badr, Yasmine Chen, Zeliang Naumov, Maxim Yao, Yantao Chen, Wenlin Kolay, Santanu Musumeci, GP Wen, Ellie Dingqiao Information Retrieval The rapidly evolving landscape of products, surfaces, policies, and regulations poses significant challenges for deploying state-of-the-art recommendation models at industry scale, primarily due to data fragmentation across domains and escalating infrastructure costs that hinder sustained quality improvements. To address this challenge, we propose Lattice, a recommendation framework centered around model space redesign that extends Multi-Domain, Multi-Objective (MDMO) learning beyond models and learning objectives. Lattice addresses these challenges through a comprehensive model space redesign that combines cross-domain knowledge sharing, data consolidation, model unification, distillation, and system optimizations to achieve significant improvements in both quality and cost-efficiency. Our deployment of Lattice at Meta has resulted in 10% revenue-driving top-line metrics gain, 11.5% user satisfaction improvement, 6% boost in conversion rate, with 20% capacity saving.
title	Meta Lattice: Model Space Redesign for Cost-Effective Industry-Scale Ads Recommendations
topic	Information Retrieval
url	https://arxiv.org/abs/2512.09200