핵심 개념
Users' interaction sequences contain noise, affecting recommendation accuracy. SSDRec proposes a three-stage framework to augment sequences and denoise effectively.
초록
SSDRec introduces a novel framework for sequence denoising in recommendation systems. The content discusses the challenges of noise in user sequences, the proposed solution of self-augmentation, and the three-stage learning paradigm of SSDRec. It includes the construction of a multi-relation graph, embedding layers, global relation encoder, self-augmentation module, and hierarchical denoising module. The content also covers model complexity analysis, evaluation metrics, datasets, baselines, and experimental results on various public datasets.
통계
"Extensive experiments on five real-world datasets demonstrate the superiority of SSDRec over state-of-the-art denoising methods."
"The proposed SSDRec model outperforms existing denoising methods by a significant margin."
인용구
"To improve reliability, we propose to augment sequences by inserting items before denoising."
"Extensive experiments on five real-world datasets demonstrate the superiority of SSDRec over state-of-the-art denoising methods."