uCTRL: Unbiased Contrastive Representation Learning via Alignment and Uniformity for Collaborative Filtering

被引:1
|
作者
Lee, Jae-woong [1 ]
Park, Seongmin [1 ]
Yoon, Mincheol [1 ]
Lee, Jongwuk [1 ]
机构
[1] Sungkyunkwan Univ, Seoul, South Korea
关键词
Alignment and uniformity; popularity bias;
D O I
10.1145/3539618.3592076
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Because implicit user feedback for the collaborative filtering (CF) models is biased toward popular items, CF models tend to yield recommendation lists with popularity bias. Previous studies have utilized inverse propensity weighting (IPW) or causal inference to mitigate this problem. However, they solely employ pointwise or pairwise loss functions and neglect to adopt a contrastive loss function for learning meaningful user and item representations. In this paper, we propose Unbiased ConTrastive Representation Learning (uCTRL), optimizing alignment and uniformity functions derived from the InfoNCE loss function for CF models. Specifically, we formulate an unbiased alignment function used in uCTRL. We also devise a novel IPW estimation method that removes the bias of both users and items. Despite its simplicity, uCTRL equipped with existing CF models consistently outperforms state-of-the-art unbiased recommender models, up to 12.22% for Recall@20 and 16.33% for NDCG@20 gains, on four benchmark datasets.
引用
收藏
页码:2456 / 2460
页数:5
相关论文
共 50 条
  • [31] Making Linear MDPs Practical via Contrastive Representation Learning
    Zhang, Tianjun
    Ren, Tongzheng
    Yang, Mengjiao
    Gonzalez, Joseph E.
    Schuurmans, Dale
    Dai, Bo
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 162, 2022,
  • [32] Unsupervised Sentence Representation via Contrastive Learning with Mixing Negatives
    Zhang, Yanzhao
    Zhang, Richong
    Mensah, Samuel
    Liu, Xudong
    Mao, Yongyi
    THIRTY-SIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FOURTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE / TWELVETH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2022, : 11730 - 11738
  • [33] Representation Learning via Adversarially-Contrastive Optimal Transport
    Cherian, Anoop
    Aeron, Shuchin
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 119, 2020, 119
  • [34] Improving graph collaborative filtering with multimodal-side-information-enriched contrastive learning
    Shan Lei
    Yuan Huanhuan
    Zhao Pengpeng
    Qu Jianfeng
    Fang Junhua
    Liu Guanfeng
    Sheng Victor S.
    Journal of Intelligent Information Systems, 2024, 62 (1) : 143 - 161
  • [35] Improving graph collaborative filtering with multimodal-side-information-enriched contrastive learning
    Shan, Lei
    Yuan, Huanhuan
    Zhao, Pengpeng
    Qu, Jianfeng
    Fang, Junhua
    Liu, Guanfeng
    Sheng, Victor S.
    JOURNAL OF INTELLIGENT INFORMATION SYSTEMS, 2024, 62 (01) : 143 - 161
  • [36] Graphs get personal: learning representation with contextual pretraining for collaborative filtering
    Shen, Tiesunlong
    Zhang, You
    Wang, Jin
    Zhang, Xuejie
    APPLIED INTELLIGENCE, 2023, 53 (24) : 30400 - 30415
  • [37] Graphs get personal: learning representation with contextual pretraining for collaborative filtering
    Tiesunlong Shen
    You Zhang
    Jin Wang
    Xuejie Zhang
    Applied Intelligence, 2023, 53 : 30416 - 30430
  • [38] Neural Binary Representation Learning for Large-Scale Collaborative Filtering
    Zhang, Yujia
    Wu, Jun
    Wang, Haishuai
    IEEE ACCESS, 2019, 7 : 60752 - 60763
  • [39] Speaker Representation Learning via Contrastive Loss with Maximal Speaker Separability
    Li, Zhe
    Mak, Man-Wai
    PROCEEDINGS OF 2022 ASIA-PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE (APSIPA ASC), 2022, : 962 - 967
  • [40] Graph Joint Representation Clustering via Penalized Graph Contrastive Learning
    Zhao, Zihua
    Wang, Rong
    Wang, Zheng
    Nie, Feiping
    Li, Xuelong
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2023, 35 (12) : 1 - 12