Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lu, Yusheng, Du, Zhaocheng, Li, Xiangyang, Jia, Pengyue, Wang, Yejing, Liu, Weiwen, Wang, Yichao, Guo, Huifeng, Tang, Ruiming, Dong, Zhenhua, Duan, Yongrui, Zhao, Xiangyu
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2408.06577
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908695702536192
author Lu, Yusheng
Du, Zhaocheng
Li, Xiangyang
Jia, Pengyue
Wang, Yejing
Liu, Weiwen
Wang, Yichao
Guo, Huifeng
Tang, Ruiming
Dong, Zhenhua
Duan, Yongrui
Zhao, Xiangyu
author_facet Lu, Yusheng
Du, Zhaocheng
Li, Xiangyang
Jia, Pengyue
Wang, Yejing
Liu, Weiwen
Wang, Yichao
Guo, Huifeng
Tang, Ruiming
Dong, Zhenhua
Duan, Yongrui
Zhao, Xiangyu
contents Large Language Models (LLMs) have exhibited significant promise in recommender systems by empowering user profiles with their extensive world knowledge and superior reasoning capabilities. However, LLMs face challenges like unstable instruction compliance, modality gaps, and high inference latency, leading to textual noise and limiting their effectiveness in recommender systems. To address these challenges, we propose UserIP-Tuning, which uses prompt-tuning to infer user profiles. It integrates the causal relationship between user profiles and behavior sequences into LLMs' prompts. It employs Expectation Maximization (EM) to infer the embedded latent profile, minimizing textual noise by fixing the prompt template. Furthermore, a profile quantization codebook bridges the modality gap by categorizing profile embeddings into collaborative IDs pre-stored for online deployment. This improves time efficiency and reduces memory usage. Experiments show that UserIP-Tuning outperforms state-of-the-art recommendation algorithms. An industry application confirms its effectiveness, robustness, and transferability. The presented solution has been deployed in Huawei AppGallery's Explore page since May 2025, serving 2 million daily active users, delivering significant improvements in real-world recommendation scenarios. The code is publicly available for replication at https://github.com/Applied-Machine-Learning-Lab/UserIP-Tuning.
format Preprint
id arxiv_https___arxiv_org_abs_2408_06577
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Prompt Tuning as User Inherent Profile Inference Machine
Lu, Yusheng
Du, Zhaocheng
Li, Xiangyang
Jia, Pengyue
Wang, Yejing
Liu, Weiwen
Wang, Yichao
Guo, Huifeng
Tang, Ruiming
Dong, Zhenhua
Duan, Yongrui
Zhao, Xiangyu
Information Retrieval
Large Language Models (LLMs) have exhibited significant promise in recommender systems by empowering user profiles with their extensive world knowledge and superior reasoning capabilities. However, LLMs face challenges like unstable instruction compliance, modality gaps, and high inference latency, leading to textual noise and limiting their effectiveness in recommender systems. To address these challenges, we propose UserIP-Tuning, which uses prompt-tuning to infer user profiles. It integrates the causal relationship between user profiles and behavior sequences into LLMs' prompts. It employs Expectation Maximization (EM) to infer the embedded latent profile, minimizing textual noise by fixing the prompt template. Furthermore, a profile quantization codebook bridges the modality gap by categorizing profile embeddings into collaborative IDs pre-stored for online deployment. This improves time efficiency and reduces memory usage. Experiments show that UserIP-Tuning outperforms state-of-the-art recommendation algorithms. An industry application confirms its effectiveness, robustness, and transferability. The presented solution has been deployed in Huawei AppGallery's Explore page since May 2025, serving 2 million daily active users, delivering significant improvements in real-world recommendation scenarios. The code is publicly available for replication at https://github.com/Applied-Machine-Learning-Lab/UserIP-Tuning.
title Prompt Tuning as User Inherent Profile Inference Machine
topic Information Retrieval
url https://arxiv.org/abs/2408.06577