I. Introduction
In recent years, knowledge graphs (KGs) have rapidly developed into a useful tool for storing information about the real world. Utilizing the triple form of (head entity, relation, tail entity), KGs can describe facts in a structured way, making connections between things clear and evident. With their strong presentation abilities, KGs are often used to perform inference prediction in numerous areas, resulting in their wide application in tasks such as question answering [1], [2], recommendation systems [3], [4] and various other natural language processing tasks [5].