chapter one

1 What is a knowledge graph?

 

This chapter covers

  • What a knowledge graph is and why it represents a paradigm shift
  • The four pillars of knowledge graphs
  • The application of knowledge graphs in multiple healthcare scenarios
  • The key technologies for knowledge graph representation

In recent years, artificial intelligence (AI) contributed to reaching unprecedented advances in many areas, helping to develop a new generation of systems to support decision making. Along with this notable progress, the current AI systems based on machine learning (ML) models have not yet reached their full potential as reliable solutions. This is due to the lack of contextual and relational information, which becomes crucial to achieving more robust, trustworthy, and explainable predictions.

Equipping ML systems with contextual information such as experiences, environmental characteristics, cultural aspects, and social norms lets us build the so-called “third wave of AI [7].” The development of such AI systems becomes strategic in situations where the impact on human life is relevant, from biotech [1] and oncology drug discovery [2] to e-commerce [3], and from intelligence [4] and law enforcement [5] to fintech [6]. In these cases, human-contextualized knowledge is not only a desirable property, but it also represents an essential requirement to empower decisions.

1.1 The knowledge graph paradigm shift

1.1.1 The four pillars of knowledge graphs

1.2 Building data-driven applications using KGs

1.2.1 360-based view for precision medicine

1.2.2 Drug discovery and development

1.2.3 Healthcare compliance management

1.2.4 Conversational AI and recommendation systems

1.2.5 What should I ask myself?

1.3 How do we teach knowledge graphs?

1.4 Knowledge graph technologies

1.5 Making graphs smarter using semantics

1.5.1 Graph vs. knowledge graph

1.5.2 Taxonomies and ontologies

1.6 Summary

1.7 References