no code implementations • 28 Sep 2023 • Chong Liu, Xiaoyang Liu, Lixin Zhang, Feng Xia, Leyu Lin
Due to the lack of supervised signals in click confidence, we first apply self-supervised learning to obtain click confidence scores via a global self-distillation method.
no code implementations • 15 Aug 2023 • Chong Liu, Xiaoyang Liu, Ruobing Xie, Lixin Zhang, Feng Xia, Leyu Lin
A powerful positive item augmentation is beneficial to address the sparsity issue, while few works could jointly consider both the accuracy and diversity of these augmented training labels.
no code implementations • 29 May 2023 • Shuai Yang, Lixin Zhang, Feng Xia, Leyu Lin
Graph-based retrieval strategies are inevitably hijacked by heavy users and popular items, leading to the convergence of candidates for users and the lack of system-level diversity.
1 code implementation • 8 Aug 2022 • Xiaoyang Liu, Chong Liu, Pinzheng Wang, Rongqin Zheng, Lixin Zhang, Leyu Lin, Zhijun Chen, Liangliang Fu
To this end, we propose a novel method that can Utilize False Negative samples for sequential Recommendation (UFNRec) to improve model performance.
no code implementations • 7 Apr 2022 • Liang Qu, Yonghong Ye, Ningzhi Tang, Lixin Zhang, Yuhui Shi, Hongzhi Yin
In order to alleviate the above issues, some works focus on automated embedding dimension search by formulating it as hyper-parameter optimization or embedding pruning problems.
2 code implementations • 13 Dec 2021 • Chong Liu, Xiaoyang Liu, Rongqin Zheng, Lixin Zhang, Xiaobo Liang, Juntao Li, Lijun Wu, Min Zhang, Leyu Lin
State-of-the-art sequential recommendation models proposed very recently combine contrastive learning techniques for obtaining high-quality user representations.