Enhanced fraud detection

• US 9,472,194 B2
• Filed: 03/21/2014
• Issued: 10/18/2016
• Est. Priority Date: 03/21/2014
• Status: Active Grant

• Current Assignee: Wells Fargo Bank NA (Wells Fargo & Company)
• Original Assignee: Wells Fargo Bank NA (Wells Fargo & Company)
• Inventors: Jones, Raymond F.
• Primary Examiner(s): ALBERTALLI, BRIAN LOUIS
• Application Number:
  US14/222,168
• Publication Number:
  US 20150269942A1
• Time in Patent Office: 942 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06Q 30/0185 Product, service or business identity fraud
• CPC Class Codes: G10L 17/02 Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction
• CPC Class Codes: G10L 17/04 Training, enrolment or model building
• CPC Class Codes: G10L 17/06 Decision making techniques; Pattern matching strategies
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces
• CPC Class Codes: G10L 17/26 Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices
• CPC Class Codes: G10L 21/0208 Noise filtering
• CPC Class Codes: H04M 15/47 Fraud detection or prevention means
• CPC Class Codes: H04M 2203/301 Management of recordings
• CPC Class Codes: H04M 3/42221 Conversation recording systems at the subscriber's set H04M1/656
• CPC Class Codes: H04M 3/5175 Call or contact centers supervision arrangements

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.

Claims

1. A system for fraud detection, comprising:

a processor that executes the following computer executable components stored in a memory;

a database component comprising a voice library that includes respective voice samples of individuals deemed to be fraudsters;

a monitoring component that;

identifies respective segments of a communication as one of a salutation segment, a verification segment, a conversation segment, or a summary segment,

isolates voice signals from segments of the respective segments of the communication, and

determines a voice match based on a comparison between the voice signals from the segments of the communication and the respective voice samples of the voice library;

an analysis component that analyzes one or more frequency responses of the voice signals from the segments of the communication to generate a frequency determination, wherein the frequency determination indicates non-stress based on a frequency of the communication being within a frequency range provided by a frequency library and indicates stress based on the frequency being outside the frequency range; and

a fraud component that identifies potential fraudulent activity based on a stress to non-stress ratio for the segments exceeding a pre-determined threshold.

2. The system of claim 1, wherein the fraud component identifies the potential fraudulent activity by applying a first weight to the voice match and a second weight to the frequency determination, the first weight is heavier than the second weight.

3. The system of claim 1, wherein the fraud component identifies the potential fraudulent activity by considering the voice match prior to considering the frequency determination.

4. The system of claim 1, wherein the monitoring component filters noise from the segments of the communication.

5. The system of claim 1, wherein the analysis component generates the frequency determination based on another comparison between the one or more frequency responses and a frequency threshold.

6. The system of claim 1, wherein the analysis component generates the frequency determination based on another comparison between the one or more frequency responses of the respective segments of the communication and one or more other frequency responses of one or more other segments of the communication.

7. The system of claim 1, wherein the fraud component provides one or more notifications to one or more parties when the potential fraudulent activity is identified.

8. The system of claim 1, wherein the monitoring component isolates the one or more voice signals for a live communication or a recorded communication.

9. The system of claim 1, wherein the one or more voice signals is from an individual.

10. The system of claim 1, wherein the one or more voice signals is from a representative of a call center.

11. A system for fraud detection, comprising:

a processing unit that executes the following computer executable components stored in a memory;

a database component comprising a pathway library that includes sets of characteristics determined to be associated with fraudulent communication;

a monitoring component that;

identifies one or more segments of a communication as a salutation segment, a verification segment, a conversation segment, or a summary segment, and

isolates one or more voice signals from one or more of the segments of the communication;

a detection component that;

determines one or more characteristics associated with the communication, and

compares the one or more characteristics associated with the communication with the sets of characteristics of the pathway library to generate a pathway match determination;

an analysis component that analyzes respective frequency responses of the one or more voice signals from the one or more segments of the communication to generate a frequency determination, wherein the frequency determination indicates non-stress based on a frequency response being within a frequency range provided by a frequency library and indicates stress based on the frequency response being outside the frequency range; and

a fraud component that identifies potential fraudulent activity based on a determination that a stress to non-stress ratio is above a pre-determined threshold.

12. The system of claim 11, wherein the one or more characteristics associated with the communication is indicative of a technology associated with the communication.

13. The system of claim 11, wherein the one or more characteristics associated with the communication is indicative of one or more artifacts associated with the communication.

14. The system of claim 11, wherein the one or more characteristics associated with the communication is indicative of noise associated with the communication.

15. The system of claim 11, wherein the communication occurs over a voice over internet protocol channel or a telecommunications channel.

16. The system of claim 11, wherein the analysis component utilizes one or more of the voice signals from the salutation segment or the conversation segment as a baseline for generating the frequency determination.

17. A system for fraud detection, comprising:

a processing unit that executes the following computer executable components stored in a memory;

a database component comprising;

a voice library comprising voice samples of individuals previously determined to be fraudsters; and

a pathway library comprising one or more sets of characteristics deemed to be associated with fraudulent communication;

a monitoring component that;

isolates voice signals from segments of a communication, wherein the segments include at least one of a salutation segment, a verification segment, a conversation segment, and a summary segment; and

compares the voice signals from the segments with the voice samples in the voice library to generate a voice match determination;

a detection component that

compares characteristics associated with the communication with the one or more sets of characteristics of the pathway library to generate a pathway match determination;

an analysis component that analyzes one or more respective frequency responses of the voice signals to generate a frequency determination that indicates respective stress to non-stress ratios for the segments; and

a fraud component that identifies potential fraudulent activity based on the respective stress to non-stress ratios exceeding a stress threshold.

18. The system of claim 17, wherein the fraud component identifies the potential fraudulent activity based on weighing the voice match determination heavier than the frequency determination.

19. The system of claim 17, wherein the fraud component identifies the potential fraudulent activity by considering the voice match determination prior to considering the frequency determination.

20. The system of claim 17, wherein the monitoring component filters noise from one or more segments of the communication.