no code implementations • 5 Aug 2023 • Yi Ren, Xu Zhao, Hongyan Tang, Shuai Li
In this paper, we propose a structural causal model-based method to address the popularity bias issue for sequential recommendation model learning.
no code implementations • 11 Apr 2023 • Yi Ren, Hongyan Tang, Jiangpeng Rong, Siwen Zhu
As pairwise learning suits well for the ranking tasks, the previously proposed unbiased pairwise learning algorithm already achieves state-of-the-art performance.
no code implementations • 8 Mar 2023 • Yi Ren, Hongyan Tang, Siwen Zhu
It is a well-known challenge to learn an unbiased ranker with biased feedback.
1 code implementation • 25 Nov 2021 • Yi Ren, Hongyan Tang, Siwen Zhu
To provide personalized high quality recommendation results, conventional systems usually train pointwise rankers to predict the absolute value of objectives and leverage a distinct shallow tower to estimate and alleviate the impact of position bias.
6 code implementations • RecSys 2020 • Hongyan Tang, Junning Liu, Ming Zhao, Xudong Gong
Moreover, through extensive experiments across SOTA MTL models, we have observed an interesting seesaw phenomenon that performance of one task is often improved by hurting the performance of some other tasks.