System and method for detecting fraud rings
A system and method may identify a fraud ring based on call or interaction data by analyzing by a computer processor interaction data including audio recordings to identify clusters of interactions which are suspected of involving fraud each cluster including the same speaker; analyzing by the computer processor the clusters, in combination with metadata associated with the interaction data, to identify fraud rings, each fraud ring describing a plurality of different speakers, each fraud ring defined by a set of speakers and a set of metadata corresponding to interactions including that speaker; and for each fraud ring, creating a relevance value defining the relative relevance of the fraud ring.
1. A method for identifying a fraud ring based on interaction data, the method comprising:
analyzing, by a computer processor, interaction data comprising audio recordings to identify clusters of interactions which are suspected of involving fraud and grouping clusters including a single speaker;
analyzing by the computer processor the clusters, in combination with metadata associated with the interaction data, to identify fraud rings, without requiring further input from a user, each fraud ring defined by a set of speakers and a set of metadata corresponding to interactions including said single speaker; and
for each fraud ring, creating a relevance value defining a relative relevance of the fraud ring.
2. The method of claim 1 wherein creating the relevance value is performed using a supervised learning method.
3. The method of claim 1 wherein each cluster is represented by a graph.
4. The method of claim 1 wherein each ring is represented by a graph.
5. The method of claim 1 wherein creating a relevance value is performed using a machine learning system.
6. The method of claim 1 wherein creating a relevance value is performed using an unsupervised learning system.
7. The method of claim 1 comprising, prior to creating a relevance value defining the relative relevance of the fraud ring:
receiving from a security officer a relevance value for a previous fraud ring, and using the received relevance value to create a relevance value defining the relative relevance of the fraud ring.
8. The method of claim 1 comprising creating a relevance value defining the relative relevance of the fraud ring by comparing the fraud ring to a fraud ring having a previously determined relevance value.
9. A system for identifying a fraud ring based on interaction data, the system comprising:
a memory; and
a processor configured to:
analyze by a computer processor interaction data comprising audio recordings to identify clusters of interactions which are suspected of involving fraud and grouping clusters including a single speaker;
analyze by the computer processor the clusters, in combination with metadata associated with the interaction data, to identify fraud rings, without requiring further input from a user, each fraud ring defined by a set of speakers and a set of metadata corresponding to interactions including said speaker; and
for each fraud ring, create a relevance value defining a relative relevance of the fraud ring.
10. The system of claim 9 wherein creating the relevance value is created using a supervised learning method.
11. The system of claim 9 wherein each cluster is represented by a graph.
12. The system of claim 9 wherein each ring is represented by a graph.
13. The system of claim 9 wherein creating a relevance value is performed using a machine learning system.
14. The system of claim 9 wherein creating a relevance value is performed using an unsupervised learning system.
15. The system of claim 9 wherein the processor is configured to, prior to creating a relevance value defining the relative relevance of the fraud ring: receive from a security officer a relevance value for a previous fraud ring, and use the received relevance value to create a relevance value defining the relative relevance of the fraud ring.
16. The system of claim 9 wherein the processor is configured to create a relevance value defining the relative relevance of the fraud ring by comparing the fraud ring to a fraud ring having a previously determined relevance value.
17. A method for identifying a fraud ring, the method comprising:
identifying clusters suspected of fraud and grouping clusters comprising audio recordings each audio recording associated with a speaker;
analyzing by the computer processor the clusters, in combination with metadata associated with the audio recordings data, to create fraud rings, each fraud ring associated with a plurality of different speakers and a set of metadata corresponding to the speakers; and
for each fraud ring, calculating a relevance score, without requiring further input from a user.
18. The method of claim 17 wherein creating the relevance score is performed using a supervised learning method.
19. The method of claim 17 wherein each cluster is represented by a graph.
20. The method of claim 17 wherein each ring is represented by a graph.