Fraud detection

Patent metadata

Field	Value
Publication number	US-9607620-B2
Application number	US-201414221590-A
Country	US
Kind code	B2
Filing date	Mar 21, 2014
Priority date	Mar 21, 2014
Publication date	Mar 28, 2017
Grant date	Mar 28, 2017

Abstract

Embodiments of techniques or systems for fraud detection are provided herein. A communication may be received where the communication includes one or more voice signals from an individual. Frequency responses associated with these voice signals may be determined and analyzed and utilized to determine whether or not potential fraudulent activity is occurring. For example, if a frequency response is greater than a frequency threshold, potential fraudulent activity may be determined. Further, frequency responses may be cross referenced with voice biometrics, voice printing, or fraud pathway detection results. In this way, voice stress or frequency responses may be utilized to build other databases related to other types of fraud detection, thereby enhancing one or more aspects of fraud detection. For example, a database may include a voice library, a pathway library, or a frequency library which include characteristics associated with fraudulent activity, thereby facilitating identification of such activity.

First claim

What is claimed is:

1. A system for fraud detection, comprising:
a processor that executes the following computer executable components stored in a memory:
a monitoring component that:
generates a voice match determination based on identification of at least two segments of a communication, wherein segments of the at least two segments comprise respective voice signals associated with an individual, and
assigns a confidence level to the voice match determination, the confidence level indicates a likelihood that the individual is a fraudster;

 an analysis component that:
analyzes respective frequency responses for each segment of the at least two segments, and
assigns a frequency response confidence level based on a determination that the respective frequency responses indicate the individual is the fraudster and based on consideration of a false positive caused by external influences; and

 a fraud component that determines fraud is occurring based on the confidence level of the voice match determination and the frequency response confidence level.

2. The system of claim 1 , wherein the at least two segments comprise a salutation segment, a verification segment, a conversation segment, or a summary segment.

3. The system of claim 1 , wherein the communication is implemented across a voice over internet protocol channel.

4. The system of claim 1 , wherein the analysis component compares the respective frequency responses based on a delta frequency threshold that represents a frequency difference between a first frequency response and a second frequency response.

5. The system of claim 1 , wherein the analysis component analyzes the respective frequency responses based on voice stress analysis.

6. The system of claim 1 , wherein the communication comprises additional voice signals associated with a second individual.

7. The system of claim 1 , the fraud component determines fraud is occurring based on the confidence level of the voice match determination being above a threshold level and the frequency response confidence level being above another threshold level.

8. The system of claim 7 , wherein the voice match determination is above the threshold level based on another determination that the individual is the same individual associated with voice signals included in a voice library.

9. The system of claim 1 , the monitoring component generates the voice match determination based on a comparison between the respective voice signals and at least one voice signal included in a voice library.

10. The system of claim 1 , wherein the analysis component cross references the respective frequency responses with at least one of a voice biometric, a voice print, or a fraud pathway detection result.

11. A method for fraud detection, comprising:
receiving, by a system comprising a processing unit, a communication that includes at least two segments, wherein segments of the at least two segments comprise respective voice signals associated with an individual;
determining, by the system, a voice match based on a comparison of the respective voice signals and at least one voice signal included in a voice library;
assigning, by the system, a confidence level to the voice match, the confidence level is based on a likelihood that the individual is a fraudster;
determining, by the system, respective frequency responses for the segments of the at least two segments;
assigning, by the system, a frequency response confidence level to the communication based on a determination that the respective frequency responses identify the individual as a fraudster and based on consideration of a false positive caused by external influences;
determining, by the system, that fraud is occurring based on the confidence level of the voice match and the frequency response confidence level; and
outputting, by the system, a fraud notification to another individual participating in the communication during a duration of the communication.

12. The method of claim 11 , further comprises comparing, by the system, the respective frequency responses based on a delta frequency threshold.

13. The method of claim 11 , further comprises analyzing, by the system, one or more of the frequency responses based on voice stress analysis.

14. The method of claim 11 , wherein the communication comprises additional voice signals associated with the another individual participating in the communication.

15. A computer-readable storage device storing executable instructions that, in response to execution, cause a system comprising a processor to perform operations, comprising:
recording a communication, wherein the communication comprises a set of segments, wherein segments of the set of segments comprise respective voice signals associated with an individual;
controlling a flow of the communication based on an order of the segments comprising introducing a segment expected to produce a stress response from the individual prior to introducing another segment expected to produce a normal response from the individual;
determining the respective voice signals are associated with an occurrence of fraud based on a comparison between the respective voice signals and voice samples included in a database;
assigning a first confidence level to the communication based on the determining;
determining respective frequency responses for segments of the set of segments;
analyzing the respective frequency responses for one or more of the segments of the communication;
assigning a second confidence level to the communication based on the analyzing; and
determining the individual is attempting fraud based on the first confidence level and the second confidence level exceeding respective threshold levels.

16. The computer-readable storage device of claim 15 , the operations further comprise analyzing one or more of the frequency responses based on voice stress analysis.

Assignees

• Jones Raymond F
• Wells Fargo Bank Na

Inventors

• Jones Raymond F

Classifications

• H04M15/47

  Fraud detection or prevention means · CPC title

• G06F21/32

  using biometric data, e.g. fingerprints, iris scans or voiceprints · CPC title

• H04M3/5175

  Call or contact centers supervision arrangements · CPC title

• G10L17/04

  Training, enrolment or model building · CPC title

• G10L17/26 Primary

  Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices · CPC title