logo
科言猫学术研究的AI总结
language
手机端手机端
  • Phone Feature微信小程序
简体中文简体中文
notice消息通知
博客博客
delete
arrow
返回

Behavior-Aware Consistent Distillation for Cold-Start Recommendation

delete2026-06-18
delete0
PRE
AI
H
Huan Gong
H
Hao Chen
L
Lijia Chen
H
Han Yang
周志博 封面图
周志博 (Zhibo Zhou)
F
Feiran Huang
K
Kai Xu
Y
Yu Yang
F
Fakhri Karray
DOI：10.1109/tkde.2026.3705082delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
冷启动场景下无交互用户和物品的推荐仍是一项基础挑战，因为内容驱动的冷启动实体与行为丰富的热实体存在显著差异。生成方法试图通过从内容特征推断嵌入向量来应对这一问题，而基于丢弃的技术则通过训练时随机屏蔽行为信号来提高鲁棒性。然而，这些方法主要关注有限的冷启动场景，未能充分捕捉冷启动与热推荐之间的差异。因此，生成模型可能产生不平衡的预测，而丢弃策略会降低热实体的推荐质量。为解决这些局限性，我们提出行为感知一致性蒸馏（BACD）框架，用于用户和物品的冷启动推荐。BACD采用行为感知的教师-学生范式，并结合行为感知的教师质量加权机制以确保可靠的监督。此外，它引入了三种一致性导向的蒸馏目标，包括评分分布一致性、排名一致性和表示分布一致性，共同缩小冷启动与热实体之间的差距。在四个基准数据集上的大量实验表明，BACD在冷启动和整体设置下均能持续提升推荐性能，并在多种骨干模型上显著优于现有最先进基线方法。
Keyword:
Cold-start recommendation
consistent distillation
content features

期刊

IEEE Transactions on Knowledge and Data Engineering 封面图
IEEE Transactions on Knowledge and Data Engineering
IF:
10.4
论文数:
6.8K
被引数:
3.2W

机构

T
tianjin university
学者数:
8.0W
论文数: 5.8W
被引数: 88
N
national university of defense technology
学者数:
5.0K
论文数: 1.5K
被引数: 0
M
Mohamed bin Zayed University of Artificial Intelligence
学者数:
128
论文数: 113
被引数: 1
T
the education university of hong kong
学者数:
601
论文数: 445
被引数: 0
C
City University of Macau
学者数:
552
论文数: 382
被引数: 2.5K
J
jinan university
学者数:
4.3W
论文数: 2.7W
被引数: 38
学者 查看更多机构
引用论文

引用论文

Rating Augmentation with Generative Adversarial Networks towards Accurate Collaborative Filtering
errThe World Wide Web Conference
IF0
err2019-05-13
err0
PREAI
errDong-Kyu Chae; Jin-Soo Kang; Sang-Wook Kim; Jaeho Choi
err分享
err收藏
RecSys Challenge 2017
errProceedings of the Eleventh ACM Conference on Recommender Systems
IF0
err2017-08-27
err0
PREAI
errFabian Abel; Yashar Deldjoo; Mehdi Elahi; Daniel Kohlsdorf
err分享
err收藏
Binarized Collaborative Filtering with Distilling Graph Convolutional Network
errProceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence
IF0
err2019-08-01
err0
errOAAI
errHaoyu Wang; Defu Lian; Yong Ge
err分享
err收藏
LARA
errProceedings of the 13th International Conference on Web Search and Data Mining
IF0
err2020-01-22
err0
PREAI
errChangfeng Sun; Han Liu; Meng Liu; Zhaochun Ren; Tian Gan; Liqiang Nie
err分享
err收藏
Few-Shot Image Recognition With Knowledge Transfer
err2019 IEEE/CVF International Conference on Computer Vision (ICCV)
IF0
err2019-10-01
err0
PREAI
errZhimao Peng; Zechao Li; Junge Zhang; Yan Li; Guo-Jun Qi; Jinhui Tang
err分享
err收藏
Aligning Distillation For Cold-start Item Recommendation
errProceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval
IF0
err2023-07-18
err0
PREAI
errFeiran Huang; Zefan Wang; Xiao Huang; Yufeng Qian; Zhetao Li; Hao Chen
err分享
err收藏
DE-RRD: A Knowledge Distillation Framework for Recommender System
errProceedings of the 29th ACM International Conference on Information & Knowledge Management
IF0
err2020-10-19
err0
errOAAI
errSeongKu Kang; Junyoung Hwang; Wonbin Kweon; Hwanjo Yu
err分享
err收藏
Collaborative Distillation for Top-N Recommendation
err2019 IEEE International Conference on Data Mining (ICDM)
IF0
err2019-11-01
err0
errOAAI
errJae-woong Lee; Minjin Choi; Jongwuk Lee; Hyunjung Shim
err分享
err收藏
Neighbor enhanced graph convolutional networks for node classification and recommendation
errKNOWLEDGE-BASED SYSTEMS
IF7.6
err2022-06-01
err32
errOAAI
errChen, Hao; Huang, Zhong; Xu, Yue; Deng, Zengde; Huang, Feiran; He, Peng; Li, Zhoujun
err分享
err收藏
Exploiting the Ground-Truth: An Adversarial Imitation Based Knowledge Distillation Approach for Event Detection
errProceedings of the AAAI Conference on Artificial Intelligence
IF0
err2019-07-17
err0
errOAAI
errJian Liu; Yubo Chen; Kang Liu
err分享
err收藏
学者 查看更多内容