Building Knowledge Graphs

Chapter 10. Pattern Detection Knowledge Graphs

Managing enterprise data as a knowledge graph provides plentiful benefits. Data is (logically) centralized, curated, and contextualized. A knowledge graph can be mined for patterns that originate from interesting business events. Those patterns give historical insight into the business, but they can also be used to look forward.

This chapter focuses on finding patterns in knowledge graphs and using those patterns to improve future outcomes. Starting with simple pattern-matching approaches, you’ll see how the patterns in your knowledge graph can be easily exploited to stop fraud and build better teams. In each of the use cases, you’ll also see how pattern detection can be augmented by graph data science to enrich the knowledge graph and surface new, valuable patterns.

Fraud Detection

Online fraud is a widespread and pernicious problem in the era of ubiquitous computer systems. These systems host some of the most critical aspects of our personal lives and are the lifeblood of modern businesses. They are tempting targets for criminals who might want to defraud individuals or organizations.

The level of that temptation is eye-watering. In the United States alone, fraud cost the productive economy $5.8 billion in 2021 according to the Federal Trade Commission, with around $1 billion being lost in the banking sector alone. Worse, the amount of money lost to fraud is increasing at a staggering 70% year on year. The picture is similar in ...