TransAct V2: Lifelong User Action Sequence Modeling on Pinterest Recommendation

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Xia, Xue, Joshi, Saurabh Vishwas, Rajesh, Kousik, Li, Kangnan, Lu, Yangyi, Pancha, Nikil, Badani, Dhruvil Deven, Xu, Jiajing, Eksombatchai, Pong
Format:	Preprint
Published:	2025
Subjects:	Information Retrieval Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

_version_	1866913872065069056
author	Xia, Xue Joshi, Saurabh Vishwas Rajesh, Kousik Li, Kangnan Lu, Yangyi Pancha, Nikil Badani, Dhruvil Deven Xu, Jiajing Eksombatchai, Pong
author_facet	Xia, Xue Joshi, Saurabh Vishwas Rajesh, Kousik Li, Kangnan Lu, Yangyi Pancha, Nikil Badani, Dhruvil Deven Xu, Jiajing Eksombatchai, Pong
contents	Modeling user action sequences has become a popular focus in industrial recommendation system research, particularly for Click-Through Rate (CTR) prediction tasks. However, industry-scale CTR models often rely on short user sequences, limiting their ability to capture long-term behavior. Additionally, these models typically lack an integrated action-prediction task within a point-wise ranking framework, reducing their predictive power. They also rarely address the infrastructure challenges involved in efficiently serving large-scale sequential models. In this paper, we introduce TransAct V2, a production model for Pinterest's Homefeed ranking system, featuring three key innovations: (1) leveraging very long user sequences to improve CTR predictions, (2) integrating a Next Action Loss function for enhanced user action forecasting, and (3) employing scalable, low-latency deployment solutions tailored to handle the computational demands of extended user action sequences.
format	Preprint
id	arxiv_https___arxiv_org_abs_2506_02267
institution	arXiv
publishDate	2025
record_format	arxiv
spellingShingle	TransAct V2: Lifelong User Action Sequence Modeling on Pinterest Recommendation Xia, Xue Joshi, Saurabh Vishwas Rajesh, Kousik Li, Kangnan Lu, Yangyi Pancha, Nikil Badani, Dhruvil Deven Xu, Jiajing Eksombatchai, Pong Information Retrieval Artificial Intelligence Modeling user action sequences has become a popular focus in industrial recommendation system research, particularly for Click-Through Rate (CTR) prediction tasks. However, industry-scale CTR models often rely on short user sequences, limiting their ability to capture long-term behavior. Additionally, these models typically lack an integrated action-prediction task within a point-wise ranking framework, reducing their predictive power. They also rarely address the infrastructure challenges involved in efficiently serving large-scale sequential models. In this paper, we introduce TransAct V2, a production model for Pinterest's Homefeed ranking system, featuring three key innovations: (1) leveraging very long user sequences to improve CTR predictions, (2) integrating a Next Action Loss function for enhanced user action forecasting, and (3) employing scalable, low-latency deployment solutions tailored to handle the computational demands of extended user action sequences.
title	TransAct V2: Lifelong User Action Sequence Modeling on Pinterest Recommendation
topic	Information Retrieval Artificial Intelligence
url	https://arxiv.org/abs/2506.02267