Knowledge Graph Completion Technology Research

In recent years, knowledge graphs have been widely used in the fields of language cognition, search and recommendation, question answering systems, etc. However, with the continuous expansion of the scale of many knowledge graphs constructed by RDF, how to solve the incompleteness of knowledge graphs has become a research hotspot. knowledge graph completion technology is used to complete the missing content in the knowledge graph, mainly including entity completion and relationship completion. This paper mainly focuses on the completion problem, and sorts out completion methods based on deep representation learning. Through the analysis of the research history and latest progress of completion technology, the practical challenges and future development directions of the technology are put forward.