Modifier Embedding for Temporal Knowledge Graph Embedding

Abstract

As a major research direction of temporal knowledge graph, the embedding and completion of temporal knowledge graph is a bridge for the application of temporal knowledge graph to various applications. Therefore, this research direction has a high position in the research of temporal knowledge graph.temporal knowledge graph embedding models are also receiving increasing attention. In this paper, we propose a modifier embedding model, and use this model to transform the quaternion of the temporal knowledge graph into a triplet. After that, we fused it with the DistMult model to obtain a new temporal knowledge graph embedding model, and conducted experiments on a dataset. Compared with existing models, our model has a good effect. After adjusting the embedding coefficients of the modified embedded model, we found that the embedding coefficients also have a certain impact on the effect of the model.