跨领域推荐方法研究综述
A review of cross-domain recommendation methods
王婷 1张悦1
作者信息
- 1. 桂林理工大学商学院,广西桂林 541004
- 折叠
摘要
近年来,随着信息技术的迅速发展的爆发性增长,这一爆发式增长推动了跨领域推荐系统的出现和发展.跨领域推荐系统的设计和实现面临着诸多挑战,包括数据异构性、领域知识融合等问题.因此,书写跨领域推荐方法的研究变得尤为重要.这些方法旨在有效地整合来自不同领域的数据和信息,同时保持推荐系统的高效性和准确性.为实现这一目标,研究者们提出了各种跨领域推荐方法,包括基于迁移学习方法、基于多任务学习的方法等跨领域推荐方法,文章将从处理步骤及优缺点梳理各跨领域推荐系统方法.
Abstract
In recent years,with the explosive growth of the rapid development of information technology,this explosive growth has promoted the emergence and development of cross-field recommendation systems.The design and implementation of cross-domain recommendation sys-tems face many challenges,including data heterogeneity and domain knowledge fusion.There-fore,the study of writing cross-field recommendation methods has become particularly impor-tant.These methods are designed to effectively integrate data and information from different domains while maintaining the efficiency and accuracy of recommender systems.In order to a-chieve this goal,researchers propose a variety of cross-domain recommendation methods,inclu-ding transfer-based learning methods,multi-task learning-based methods and other cross-domain recommendation methods.
关键词
迁移学习/多任务学习/共享表示学习/迁移策略学习/元学习/混合方法学习/基于主题模型和知识图像学习Key words
transfer learning/multi-task learning/shared representation/migration strategy/me-ta-learning/Hybrid approach/Subject-based model and knowledge-based image learning引用本文复制引用
出版年
2024
NETL
NSTL
万方数据