事例

Connecting Claims to Identify Fraud Networks

読むのに3分Consilience Team

これでできることInvestigators gain a connected view of relationships and supporting evidence that may otherwise be difficult to identify across individual claims.

Insurance fraud can be difficult to detect because individual claims may appear legitimate when reviewed separately. Suspicious behavior may only become visible when relationships between different claims, people, businesses, addresses, or service providers are analyzed together.

See the connections hidden between individual claims

For example, several unrelated claimants may repeatedly use the same repair shop, doctor, lawyer, address, or phone number. Those connections may indicate an organized fraud network.

Consilience could connect these entities in a knowledge graph and allow investigators to analyze patterns across many different claims.

Questions investigators could ask

Investigators could ask:

  • “Which people or businesses repeatedly appear across suspicious claims?”
  • “Which claims share the same repair shop, doctor, lawyer, or address?”
  • “Show the relationships between these claims.”
  • “Are there unusual clusters inside the claims network?”

Support existing fraud-detection systems with connected evidence

Consilience would not necessarily replace existing fraud-detection systems. Instead, it could provide investigators with a connected view of relationships and supporting evidence that may otherwise be difficult to identify.

Connecting Claims to Identify Fraud Networks · Consilience