基于正则化流的神经过程用于少样本知识图谱完形填空（英文版）

英文标题：Normalizing Flow-based Neural Process for Few-Shot Knowledge Graph Completion中文摘要：本文提出了一种基于正则流的神经过程来实现几个少样本知识图谱补全，可以处理复杂关系并估计预测的不确定性，并且引入注意力关系路径图神经网络来捕捉知识图谱中的路径信息，并在三个公共数据集上显着优于现有方法并实现了最先进的性能。英文摘要：Knowledge graphs (KGs), as a structured form of knowledge representation,have been widely applied in the real world. Recently, few-shot knowledge graphcompletion (FKGC), which aims to predict missing facts for unseen relationswith few-shot associated facts, has attracted increasing attention from