Learning Recommender Systems with Soft Target: A Decoupled Perspective

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Hao, Cheng, Mingyue, Liu, Qi, Luo, Yucong, Li, Rui, Chen, Enhong
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909342982209536
author Zhang, Hao
Cheng, Mingyue
Liu, Qi
Luo, Yucong
Li, Rui
Chen, Enhong
author_facet Zhang, Hao
Cheng, Mingyue
Liu, Qi
Luo, Yucong
Li, Rui
Chen, Enhong
contents Learning recommender systems with multi-class optimization objective is a prevalent setting in recommendation. However, as observed user feedback often accounts for a tiny fraction of the entire item pool, the standard Softmax loss tends to ignore the difference between potential positive feedback and truly negative feedback. To address this challenge, we propose a novel decoupled soft label optimization framework to consider the objectives as two aspects by leveraging soft labels, including target confidence and the latent interest distribution of non-target items. Futhermore, based on our carefully theoretical analysis, we design a decoupled loss function to flexibly adjust the importance of these two aspects. To maximize the performance of the proposed method, we additionally present a sensible soft-label generation algorithm that models a label propagation algorithm to explore users' latent interests in unobserved feedback via neighbors. We conduct extensive experiments on various recommendation system models and public datasets, the results demonstrate the effectiveness and generality of the proposed method.
format Preprint
id arxiv_https___arxiv_org_abs_2410_06536
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Learning Recommender Systems with Soft Target: A Decoupled Perspective
Zhang, Hao
Cheng, Mingyue
Liu, Qi
Luo, Yucong
Li, Rui
Chen, Enhong
Information Retrieval
Learning recommender systems with multi-class optimization objective is a prevalent setting in recommendation. However, as observed user feedback often accounts for a tiny fraction of the entire item pool, the standard Softmax loss tends to ignore the difference between potential positive feedback and truly negative feedback. To address this challenge, we propose a novel decoupled soft label optimization framework to consider the objectives as two aspects by leveraging soft labels, including target confidence and the latent interest distribution of non-target items. Futhermore, based on our carefully theoretical analysis, we design a decoupled loss function to flexibly adjust the importance of these two aspects. To maximize the performance of the proposed method, we additionally present a sensible soft-label generation algorithm that models a label propagation algorithm to explore users' latent interests in unobserved feedback via neighbors. We conduct extensive experiments on various recommendation system models and public datasets, the results demonstrate the effectiveness and generality of the proposed method.
title Learning Recommender Systems with Soft Target: A Decoupled Perspective
topic Information Retrieval
url https://arxiv.org/abs/2410.06536