HIT Model: A Hierarchical Interaction-Enhanced Two-Tower Model for Pre-Ranking Systems

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yang, Haoqiang, Yuan, Congde, Bai, Kun, Guo, Mengzhuo, Yang, Wei, Zhou, Chao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913970860851200
author Yang, Haoqiang
Yuan, Congde
Bai, Kun
Guo, Mengzhuo
Yang, Wei
Zhou, Chao
author_facet Yang, Haoqiang
Yuan, Congde
Bai, Kun
Guo, Mengzhuo
Yang, Wei
Zhou, Chao
contents Online display advertising platforms rely on pre-ranking systems to efficiently filter and prioritize candidate ads from large corpora, balancing relevance to users with strict computational constraints. The prevailing two-tower architecture, though highly efficient due to its decoupled design and pre-caching, suffers from cross-domain interaction and coarse similarity metrics, undermining its capacity to model complex user-ad relationships. In this study, we propose the Hierarchical Interaction-Enhanced Two-Tower (HIT) model, a new architecture that augments the two-tower paradigm with two key components: $\textit{generators}$ that pre-generate holistic vectors incorporating coarse-grained user-ad interactions through a dual-generator framework with a cosine-similarity-based generation loss as the training objective, and $\textit{multi-head representers}$ that project embeddings into multiple latent subspaces to capture fine-grained, multi-faceted user interests and multi-dimensional ad attributes. This design enhances modeling effectiveness without compromising inference efficiency. Extensive experiments on public datasets and large-scale online A/B testing on Tencent's advertising platform demonstrate that HIT significantly outperforms several baselines in relevance metrics, yielding a $1.66\%$ increase in Gross Merchandise Volume and a $1.55\%$ improvement in Return on Investment, alongside similar serving latency to the vanilla two-tower models. The HIT model has been successfully deployed in Tencent's online display advertising system, serving billions of impressions daily. The code is available at https://github.com/HarveyYang123/HIT_model.
format Preprint
id arxiv_https___arxiv_org_abs_2505_19849
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HIT Model: A Hierarchical Interaction-Enhanced Two-Tower Model for Pre-Ranking Systems
Yang, Haoqiang
Yuan, Congde
Bai, Kun
Guo, Mengzhuo
Yang, Wei
Zhou, Chao
Information Retrieval
Online display advertising platforms rely on pre-ranking systems to efficiently filter and prioritize candidate ads from large corpora, balancing relevance to users with strict computational constraints. The prevailing two-tower architecture, though highly efficient due to its decoupled design and pre-caching, suffers from cross-domain interaction and coarse similarity metrics, undermining its capacity to model complex user-ad relationships. In this study, we propose the Hierarchical Interaction-Enhanced Two-Tower (HIT) model, a new architecture that augments the two-tower paradigm with two key components: $\textit{generators}$ that pre-generate holistic vectors incorporating coarse-grained user-ad interactions through a dual-generator framework with a cosine-similarity-based generation loss as the training objective, and $\textit{multi-head representers}$ that project embeddings into multiple latent subspaces to capture fine-grained, multi-faceted user interests and multi-dimensional ad attributes. This design enhances modeling effectiveness without compromising inference efficiency. Extensive experiments on public datasets and large-scale online A/B testing on Tencent's advertising platform demonstrate that HIT significantly outperforms several baselines in relevance metrics, yielding a $1.66\%$ increase in Gross Merchandise Volume and a $1.55\%$ improvement in Return on Investment, alongside similar serving latency to the vanilla two-tower models. The HIT model has been successfully deployed in Tencent's online display advertising system, serving billions of impressions daily. The code is available at https://github.com/HarveyYang123/HIT_model.
title HIT Model: A Hierarchical Interaction-Enhanced Two-Tower Model for Pre-Ranking Systems
topic Information Retrieval
url https://arxiv.org/abs/2505.19849