Learning Explicit User Interest Boundary for Recommendation

1 code implementation • 22 Nov 2021 • Jianhuan Zhuo, Qiannan Zhu, Yinliang Yue, Yuhong Zhao

The core objective of modelling recommender systems from implicit feedback is to maximize the positive sample score $s_p$ and minimize the negative sample score $s_n$, which can usually be summarized into two paradigms: the pointwise and the pairwise.

Recommendation Systems