Practice vocabulary for enterprise knowledge graph use cases: product catalog graphs, fraud detection, graph federation, golden record, and entity resolution at scale.
0 / 5 completed
1 / 5
What is a 'golden record' in the context of enterprise knowledge graphs?
A golden record is the master entity record created by resolving duplicates across source systems. For example, the same customer may exist in CRM, billing, and support systems with slightly different names — the golden record merges them into one canonical identity in the knowledge graph.
2 / 5
How is a knowledge graph used in fraud detection?
Graph-based fraud detection exploits the connected nature of fraud. A graph can reveal that ten seemingly unrelated accounts share a phone number, or that money flows through a chain of accounts before withdrawal — patterns invisible in flat tabular data.
3 / 5
What does 'graph federation' mean in an enterprise architecture context?
Graph federation allows an enterprise to maintain separate domain graphs (e.g., product, customer, supply chain) while enabling cross-domain queries. A federated layer routes queries to the appropriate graphs and merges results transparently.
4 / 5
What is 'entity resolution at scale' and why is it challenging?
Entity resolution (deduplication, record linkage) must determine which records across systems represent the same entity. At scale, comparing every pair is computationally infeasible, so techniques like blocking, embedding-based similarity, and probabilistic matching are used.
5 / 5
Which sentence correctly describes using a knowledge graph 'as a platform'?
'Knowledge graph as a platform' means the graph becomes a central, reusable infrastructure component. Instead of each product team building its own data model, they all read from and contribute to the shared graph, reducing duplication and enabling cross-domain insights.