Research on Prediction of Link Embedding in Maritime Knowledge Graph

The knowledge graph is essentially a multi-relational network, which uses a structured method to store the knowledge system of the corresponding relationships between entities in the real world. It is a structured representation of the real world and brings new technical means to the knowledge representation of the maritime world. However, the data is often incomplete, missing many nodes or links, because they may be only a small part of all based on credible facts. We can solve this problem by predicting missing links. Recent studies on multi-hop KGQA try to use relevant external text to deal with KG sparsity, this method still has relatively large limitations. In another study, a KG embedding method has been proposed to reduce KG sparsity by performing missing link prediction.