Knowledge Graph Representation: Capturing World Knowledge with Machine Learning
The ability to reason about the world in terms of entities and relationships between them is an important aspect of human and machine cognition. Learning to mathematically represent entities and relations in a way that captures their meaning – the objective of knowledge graph representation models – is fundamental to the advancement of machine learning and natural language processing research. R
epresentations learned from knowledge graphs (i.e. large collections of real world facts) are used for various tasks, such as question answering, information retrieval, fact checking, and inferring new facts. In this talk, I will give an introduction to representation learning from knowledge graphs, followed by an overview of recent advances in the field, with the main focus on linear factorisation models.