Integrating Item Relevance in Training Loss for Sequential Recommender Systems

被引:3
|
作者
Bacciu, Andrea [1 ]
Siciliano, Federico [1 ]
Tonellotto, Nicola [2 ]
Silvestri, Fabrizio [1 ]
机构
[1] Sapienza Univ Rome, Rome, Italy
[2] Univ Pisa, Pisa, Italy
来源
PROCEEDINGS OF THE 17TH ACM CONFERENCE ON RECOMMENDER SYSTEMS, RECSYS 2023 | 2023年
关键词
Recommender systems; Sequential recommendation; Item relevance;
D O I
10.1145/3604915.3610643
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Sequential Recommender Systems (SRSs) are a popular type of recommender system that leverages user history to predict the next item of interest. However, the presence of noise in user interactions, stemming from account sharing, inconsistent preferences, or accidental clicks, can significantly impact the robustness and performance of SRSs, particularly when the entire item set to be predicted is noisy. This situation is more prevalent when only one item is used to train and evaluate the SRSs. To tackle this challenge, we propose a novel approach that addresses the issue of noise in SRSs. First, we propose a sequential multi-relevant future items training objective, leveraging a loss function aware of item relevance, thereby enhancing their robustness against noise in the training data. Additionally, to mitigate the impact of noise at evaluation time, we propose multi-relevant future items evaluation (MRFI-evaluation), aiming to improve overall performance. Our relevance-aware models obtain an improvement of 1.58% of NDCG@10 and 0.96% in terms of HR@10 in the traditional evaluation protocol, the one which utilizes one relevant future item. In the MRFI-evaluation protocol, using multiple future items, the improvement is 2.82% of NDCG@10 and 0.64% of HR@10 w.r.t the best baseline model.
引用
收藏
页码:1114 / 1119
页数:6
相关论文
共 50 条
  • [1] Integrating Item Based Stereotypes in Recommender Systems
    ALRossais, Nourah A.
    PROCEEDINGS OF THE 26TH CONFERENCE ON USER MODELING, ADAPTATION AND PERSONALIZATION (UMAP'18), 2018, : 265 - 268
  • [2] Integrating optimized item selection with active learning for continuous exploration in recommender systems
    Kadioglu, Serdar
    Kleynhans, Bernard
    Wang, Xin
    ANNALS OF MATHEMATICS AND ARTIFICIAL INTELLIGENCE, 2024,
  • [3] A Reproducible Analysis of Sequential Recommender Systems
    Betello, Filippo
    Purificato, Antonio
    Siciliano, Federico
    Trappolini, Giovanni
    Bacciu, Andrea
    Tonellotto, Nicola
    Silvestri, Fabrizio
    IEEE ACCESS, 2025, 13 : 5762 - 5772
  • [4] Investigating the Robustness of Sequential Recommender Systems Against Training Data Perturbations
    Betello, Filippo
    Siciliano, Federico
    Mishra, Pushkar
    Silvestri, Fabrizio
    ADVANCES IN INFORMATION RETRIEVAL, ECIR 2024, PT II, 2024, 14609 : 205 - 220
  • [5] Disentangled Item Representation for Recommender Systems
    Cui, Zeyu
    Yu, Feng
    Wu, Shu
    Liu, Qiang
    Wang, Liang
    ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY, 2021, 12 (02)
  • [6] TRec: Sequential Recommender Based On Latent Item Trend Information
    Tao, Ye
    Wang, Can
    Yao, Lina
    Li, Weimin
    Yu, Yonghong
    2020 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2020,
  • [7] Bloofi Representation for Item/User in Recommender Systems
    Farahi, Zahra
    Moeini, Ali
    Kamandi, Ali
    2019 5TH INTERNATIONAL CONFERENCE ON WEB RESEARCH (ICWR), 2019, : 67 - 73
  • [8] Efficient and Fair Item Coverage in Recommender Systems
    Koutsopoulos, Iordanis
    Halkidi, Maria
    2018 16TH IEEE INT CONF ON DEPENDABLE, AUTONOM AND SECURE COMP, 16TH IEEE INT CONF ON PERVAS INTELLIGENCE AND COMP, 4TH IEEE INT CONF ON BIG DATA INTELLIGENCE AND COMP, 3RD IEEE CYBER SCI AND TECHNOL CONGRESS (DASC/PICOM/DATACOM/CYBERSCITECH), 2018, : 912 - 918
  • [9] Sequential Attack Detection in Recommender Systems
    Aktukmak, Mehmet
    Yilmaz, Yasin
    Uysal, Ismail
    IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY, 2021, 16 : 3285 - 3298
  • [10] Effect of Item Representation and Item Comparison Models on Metrics for Surprise in Recommender Systems
    de Lima, Andre Paulino
    Peres, Sarajane Marques
    PROCEEDINGS OF THE 21ST INTERNATIONAL CONFERENCE ON ENTERPRISE INFORMATION SYSTEMS (ICEIS), VOL 1, 2019, : 513 - 524