论文标题
GIDN：用于高效链路预测的轻量级构造扩散网络
GIDN: A Lightweight Graph Inception Diffusion Network for High-efficient Link Prediction
论文作者
论文摘要
在本文中，我们提出了一个图形扩散网络（GIDN）模型。该模型概括了不同特征空间中的图形扩散，并使用Inception模块来避免由复杂的网络结构引起的大量计算。我们在开放图基准（OGB）数据集上评估GIDN模型，比OGBL-Collab数据集的AGDN高11％。
In this paper, we propose a Graph Inception Diffusion Networks(GIDN) model. This model generalizes graph diffusion in different feature spaces, and uses the inception module to avoid the large amount of computations caused by complex network structures. We evaluate GIDN model on Open Graph Benchmark(OGB) datasets, reached an 11% higher performance than AGDN on ogbl-collab dataset.
