PASSLEAF: A Pool-bAsed Semi-Supervised LEArning Framework for Uncertain Knowledge Graph Embedding

In this paper, we study the problem of embedding uncertain knowledge graphs, where each relation between entities is associated with a confidence score. Observing existing methods may discard uncertainty information, only incorporate specific type score function, or cause many false-negative samples in training, propose PASSLEAF framework to solve above issues. consists two parts, one model that can different types scoring functions predict scores and other semi-supervised learning by exploiting both positive negative estimated scores. Furthermore, leverages sample pool as relay generated further augment learning. Experiment results show our proposed learn better terms having higher accuracy prediction tail entity prediction.