论文标题:Multi-Level Graph Contrastive Learning论文作者:Pengpeng Shao, Tong Liu, Dawei Zhang, J. Tao, Feihu Che, Guohua Yang论文来源:2021, Neurocomputing论文地址:download论文代码:download 1 Introduction本文贡献:提出多层次图对比学习框架:联合节点级和图级对比学习; 引入KNN 图提取语义信息;...
Graph Encoder Layer 本文采用特定图编码器层提取节点特征,利用LightGCN的有效性和轻量级架构提取协同信息和共同作用信息。\begin{split} \mathbf{X}^{(l)} = \text{GraphEncoder}(\mathbf{X}, \mathbf{A}) = \mathbf{D^{-\frac{1}{2}} A D^{-\frac{1}{2}}} \mathbf{X}^{(l-1)}, \\ \end...
To fill this gap, we propose a Multi-Level Knowledge Graph Contrastive Learning framework (ML-KGCL) to introduce CL into the KG-based recommendation. ML-KGCL makes the CL task more compatible with the recommendation task while mitigating the long-tail issue by performing fine-grained node ...
a multi-source medical knowledge augmented medication prediction network, which incorporates a multi-level graph contrastive learning framework. To the best of our knowledge, our model is the first to capture valuable relations between medical codes and augment their representations using a cascaded...
X(l)=GraphEncoder(X,A)=D−12AD−12X(l−1) 3. 方法 如图2 所示为方法框架图。它包括四个主要部分: 1)图构建层。它通过聚合用户行为序列来构造用户-商品、用户-用户和商品-商品图; 2)兴趣级别的对比学习层。它首先从用户行为序列中学习用户当前的兴趣,从用户-商品图中学习用户的总体兴趣,然后...
However, the current graph contrastive learning framework has two limitations. First, the augmentations are designed for general graphs and thus may not be suitable or powerful enough for certain domains. Second, the contrastive scheme only learns representations that are invariant to local ...
the diversity of entity representations in different contexts. We consider that the schema of KG is beneficial for preserving the consistency of entities across contexts, and we propose a novelschema-augmentedmulti-level contrastivelearning framework (SMiLE😊) to conduct knowledge graph link prediction...
The source code for "Multi-level Cross-view Contrastive Learning for Knowledge-aware Recommender System". - CCIIPLab/MCCLK
It is worth noting that we also introduce a cross-signal contrastive learning layer to allow guidance informative exchange between the two sub-modules. To the best of our knowledge, this is the first model that can comprehensively explore the benefits of both “soft” and “hard” denoising ...
Graph convolutional network with triplet attention learning for person re-identification 2022, Information Sciences Show abstract Person re-identification with part prediction alignment 2021, Computer Vision and Image Understanding Show abstract ReID-DeePNet: A Hybrid Deep Learning System for Person Re-Identi...