Probabilistic Metric Learning with Adaptive Margin for Top-K Recommendation

被引:26
|
作者
Ma, Chen [1 ,2 ]
Ma, Liheng [1 ,2 ]
Zhang, Yingxue [2 ]
Tang, Ruiming [3 ]
Liu, Xue [1 ]
Coates, Mark [1 ]
机构
[1] McGill Univ, Montreal, PQ, Canada
[2] Huawei Noahs Ark Lab Montreal, Montreal, PQ, Canada
[3] Huawei Noahs Ark Lab, Shenzhen, Peoples R China
关键词
D O I
10.1145/3394486.3403147
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Personalized recommender systems are playing an increasingly important role as more content and services become available and users struggle to identify what might interest them. Although matrix factorization and deep learning based methods have proved effective in user preference modeling, they violate the triangle inequality and fail to capture fine-grained preference information. To tackle this, we develop a distance-based recommendation model with several novel aspects: (i) each user and item are parameterized by Gaussian distributions to capture the learning uncertainties; (ii) an adaptive margin generation scheme is proposed to generate the margins regarding different training triplets; (iii) explicit user-user/item-item similarity modeling is incorporated in the objective function. The Wasserstein distance is employed to determine preferences because it obeys the triangle inequality and can measure the distance between probabilistic distributions. Via a comparison using five real-world datasets with state-of-the-art methods, the proposed model outperforms the best existing models by 4-22% in terms of recall@K on Top-K recommendation.
引用
收藏
页码:1036 / 1044
页数:9
相关论文
共 50 条
  • [1] Symmetric Metric Learning with Adaptive Margin for Recommendation
    Li, Mingming
    Zhang, Shuai
    Zhu, Fuqing
    Qia, Wanhui
    Zang, Liangjun
    Han, Jizhong
    Hu, Songlin
    [J]. THIRTY-FOURTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THE THIRTY-SECOND INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE AND THE TENTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2020, 34 : 4634 - 4641
  • [2] Differentiable Ranking Metric Using Relaxed Sorting for Top-K Recommendation
    Lee, Hyunsung
    Cho, Sangwoo
    Jang, Yeongjae
    Kim, Jaekwang
    Woo, Honguk
    [J]. IEEE ACCESS, 2021, 9 : 114649 - 114658
  • [3] Differentiable Ranking Metric Using Relaxed Sorting for Top-K Recommendation
    Lee, Hyunsung
    Cho, Sangwoo
    Jang, Yeongjae
    Kim, Jaekwang
    Woo, Honguk
    [J]. IEEE Access, 2021, 9 : 114649 - 114658
  • [4] Transfer learning from rating prediction to Top-k recommendation
    Ye, Fan
    Lu, Xiaobo
    Li, Hongwei
    Chen, Zhenyu
    [J]. PLOS ONE, 2024, 19 (03):
  • [5] Location-aware online learning for top-k recommendation
    Palovics, Robert
    Szalai, Peter
    Pap, Julia
    Frigo, Erzsebet
    Kocsis, Levente
    Benczur, Andras A.
    [J]. PERVASIVE AND MOBILE COMPUTING, 2017, 38 : 490 - 504
  • [6] On Sampling Top-K Recommendation Evaluation
    Li, Dong
    Jin, Ruoming
    Gao, Jing
    Liu, Zhi
    [J]. KDD '20: PROCEEDINGS OF THE 26TH ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY & DATA MINING, 2020, : 2114 - 2124
  • [7] Probabilistic Reverse Top-k Queries
    Jin, Cheqing
    Zhang, Rong
    Kang, Qiangqiang
    Zhang, Zhao
    Zhou, Aoying
    [J]. DATABASE SYSTEMS FOR ADVANCED APPLICATIONS, DASFAA 2014, PT I, 2014, 8421 : 406 - 419
  • [8] Probabilistic Reverse Top-k Query on Probabilistic Data
    Trieu Minh Nhut Le
    Cao, Jinli
    [J]. DATABASES THEORY AND APPLICATIONS, ADC 2023, 2024, 14386 : 30 - 43
  • [9] Top-k Link Recommendation in Social Networks
    Song, Dongjin
    Meyer, David A.
    Tao, Dacheng
    [J]. 2015 IEEE INTERNATIONAL CONFERENCE ON DATA MINING (ICDM), 2015, : 389 - 398
  • [10] Top-k spatial joins of probabilistic objects
    Ljosa, Vebjorn
    Singh, Ambuj K.
    [J]. 2008 IEEE 24TH INTERNATIONAL CONFERENCE ON DATA ENGINEERING, VOLS 1-3, 2008, : 566 - +