A knowledge graph (KG) is a collection of real-world facts represented as triples. Each triple $$(h, r, t)$$ denotes a relation $$r$$ between a head entity $$h$$ and a tail entity $$t$$. Formally, a KG can be represented as a directed multi-relational graph, where nodes correspond to entities and edges encode the relations connecting them.

University of Michigan - Ann Arbor

Google

A survey of embedding models of entities and relationships for knowledge graph completion

because knowledge graphs are typically incomplete, it is useful to perform knowledge graph completion or link prediction, i.e. predict whether a relationship not in the knowledge graph is likely to be true.

When to use the Knowledge Graphs


This paper serves as a comprehensive survey of embedding models of entities and relationships for knowledge graph completion, summarizing up-to-date experimental results on standard benchmark datasets and pointing out potential future research directions.

Purpose of this survey

Future research directions in knowledge graph completion include: (1) combining logical rules containing rich background information and knowledge graph (KG) triples in a unified framework (e.g., jointly embedding KGs and logical rules); (2) exploring open-world KG completion models to connect unseen entities, addressing the limitation of closed-world assumptions in recent models; and (3) investigating efficient approaches applicable to large-scale KGs with millions of entities and relations.

Future Research Directions in Knowledge Graph Completion

The Path Ranking Algorithm (PRA) is a random walk inference technique for knowledge graph completion. It predicts a new relationship between two entities by estimating the probability of an unseen triple as a combination of weighted random walks that follow different paths linking the head entity and the tail entity in the knowledge graph.

Learn Before

Related