Objectives

At the end of the course, you will be able to design and apply Knowledge Graphs (KGs) to solve a range of problems in modern information systems. In particular, you will be able to make educated assessments of using logic- or machine learning (ML)-based representations of KGs, and apply them in practice. You will be able to design scalable systems that utilise KGs, and gain practical experience in developing Knowledge Graph-based applications.

Contents

The course will cover three main areas:

• Representations of Knowledge Graphs
  (logic- and machine learning (ML)-based)
• Systems for Knowledge Graphs
  (scalability and reasoning)
• Applications of Knowledge Graphs
  (real-world enterprise artificial intelligence)

Remark: A fully detailed, graphical, overview can be found at https://knowledgegraph.science/PRO-KGS

In more detail, the course will cover the following topic areas. In terms of representations we will cover Knowledge Graph Embeddings, a widely applied, large family of machine learning (ML) models, Logical Knowledge in KGs, a highly expressive, diverse family of logical models, and Graph Neural Networks, which are ML-methods that use the KG structure as the basis of a neural network. We will also take a glimpse into Data Models for Knowledge Graphs.

In terms of systems we will cover Architectures, i.e., the big picture of buiding IT architectures for KGs, and its close companion Scalable Reasoning methods for making use of the knowledge in the KG. We will also discuss KG Creation, including methods such as Inductive Logic Programming (ILP), and KG Evolution, i.e., how to update, correct, and complete a KG.

In terms of applications we will cover the diverse Real-World Applications, such as banking software, medical informatics, supply chains, etc., and take a particular focus on Financial KGs with concrete applications in finance and economics. We will also take a glimpse at Services that can be provided provide based on KGs, as well as Connections between KGs, Artificial Intelligence (AI), Machine Learning (ML) and Data Science.

Requirements

Some familiarity with databases would be useful; Database Design would be an ideal preparation.