Systems and methods for detecting bust out fraud using credit data

• US 7,991,689 B1
• Filed: 07/23/2008
• Issued: 08/02/2011
• Est. Priority Date: 07/23/2008
• Status: Active Grant

• Current Assignee: Experian Information Solutions Incorporated (Experian plc)
• Original Assignee: Experian Information Solutions Incorporated (Experian plc)
• Inventors: Brunzell, Hakan Olof, Wong, Tak Wun, Caron, Arielle Renee
• Primary Examiner(s): CAMPEN, KELLY SCAGGS
• Application Number:
  US12/220,320
• Time in Patent Office: 1,105 Days
• Field of Search: 705/38, 705/35, 705/39, 705/10, 705/26, 235/379, 235/380
• US Class Current: 705/38
• CPC Class Codes: G06Q 40/02 Banking, e.g. interest calculation or account maintenance credit or loans G06Q40/03
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

Systems and methods are for, for example, predicting bust out fraud using credit bureau data. In one embodiment, credit bureau scoring models are created using credit bureau data to detect bust out fraud. The credit bureau scoring models may be then applied to consumer data to determine whether a consumer is involved in bust out fraud.

Claims

1. A computer implemented method of developing a data model for automatically identifying bust out fraud records comprising:

accessing computer-executable instructions from at least one computer-readable storage medium; and

executing the computer-executable instructions, thereby causing computer hardware comprising at least one computer processor to perform operations comprising;

receiving a set of only credit bureau data;

analyzing the set of only credit bureau data to develop a credit bureau bust out model that predicts the propensity of a consumer to be engaged in bust out fraud;

receiving a selected consumer record related to the consumer;

applying the credit bureau bust out filter in real time to the selected consumer record; and

generating a credit bureau bust out score for the selected consumer record, the credit bureau bust out score indicating whether the consumer is engaged in bust out fraud,

wherein bust out fraud involves opening one or more credit accounts across one or more credit issuers and initially exhibiting normal consumer behavior followed by a large number of purchases, cash advances, or other uses of credit before abandonment of the one or more accounts.

2. The computer implemented method of claim 1 further comprising:

identifying a first set of credit bureau data from a credit bureau database that are defined as bust out records related to consumers that are or were involved in bust out fraud; and

identifying a second set of credit bureau data from the credit bureau database that are defined as non-fraud records;

wherein the set of only credit bureau data comprises the first set and the second set of credit bureau data from the credit bureau database.

3. The computer implemented method of claim 2, wherein the second set of records includes approximately 20 to 13000 times more records than the first set of records.

4. The computer implemented method of claim 1 further comprising storing the credit bureau bust out filter in a data repository.

5. The computer implemented method of claim 4, further comprising conducting additional analysis to develop the credit bureau bust out filter.

6. The computer implemented method of claim 1, further comprising periodically applying the credit bureau bust out filter to a set of consumer records to generate credit bureau bust out scores for each of the set of consumer records related to a set of consumers, the credit bureau bust out scores indicating whether each of the consumers is engaged in bust out fraud.

7. A non-transitory storage medium having a computer program stored thereon for causing a suitably programmed system to process computer-program code by performing a method when the computer program is executed on the suitably programmed system, the method comprising:

receiving a set of only credit bureau data;

analyzing the set of only credit bureau data to develop a credit bureau bust out filter that predicts the propensity of a consumer to be engaged in bust out fraud;

receiving a selected consumer record related to the consumer;

applying the credit bureau bust out filter in real time to the selected consumer record; and

generating a credit bureau bust out score for the selected consumer record, the credit bureau bust out score indicating whether the consumer is engaged in bust out fraud,

wherein bust out fraud involves opening one or more credit accounts across one or more credit issuers and initially exhibiting normal consumer behavior followed by a large number of purchases, cash advances, or other uses of credit before abandonment of the one or more accounts.

8. The storage medium of claim 7, wherein the credit bureau data is received from one or more data sources.

9. A bust out fraud detection system, the system comprising:

a processor configured to run software modules;

a data storage device storing a plurality of consumer records, the data storage device in electronic communication with the processor; and

a bust out module configured to;

identify a subset of the plurality of consumer records from the data storage device;

receive a credit bureau bust out model from a storage repository, the credit bureau bust out model predicting which consumer records are likely involved in bust out fraud and the credit bureau bust out model created based on only credit bureau data;

apply the credit bureau bust out model to each of the subset of the plurality of consumer records to generate a credit bureau bust out score for each of the subset of the plurality of consumer records; and

store in a storage repository the credit bureau bust out score associated with the subset of the plurality of consumer records; and

the processor able to run the bust out module,

wherein bust out fraud exhibits a consumer opening one or more credit accounts across one or more credit issuers and initially exhibiting normal consumer behavior followed by a large number of purchases, cash advances, or other uses of credit before abandonment of the one or more accounts.

10. The bust out fraud detecting system of claim 9, wherein the plurality of consumer records is received in real time.

11. The bust out fraud detection system of claim 9, wherein the plurality of consumer records relate to prospective consumers that may be approved for credit.

12. The bust out fraud detection system of claim 9, wherein the plurality of consumer records is received in a batch.

13. The bust out fraud detection system of claim 9, wherein the plurality of consumer records represent existing consumer accounts.

14. The bust out fraud detection system of claim 9, wherein the credit bureau bust out model predicts fraud one to three months in advance.

15. The bust out fraud detection system of claim 9, wherein the credit bureau data is accessible from one or more data sources.

16. The bust out fraud detection system of claim 15, further comprising conduct additional analysis to develop the credit bureau bust out model.

17. A computer implemented method for generating scores that indicate bust out fraud comprising:

accessing computer-executable instructions from at least one computer-readable storage medium; and

executing the computer-executable instructions, thereby causing computer hardware comprising at least one computer processor to perform operations comprising;

identifying a plurality of consumer records;

receiving a bust out model from a storage repository, the bust out model created using substantially only credit bureau data;

applying the bust out model to each of the plurality consumer records to generate a bust out score for each of the plurality of consumer records; and

storing in a data storage the bust out score associated with each of the plurality of consumer records,

wherein bust out fraud exhibits a consumer opening one or more credit accounts across one or more credit issuers and initially exhibiting normal consumer behavior followed by a large number of purchases, cash advances, or other uses of credit before abandonment of the one or more accounts.

18. The computer implemented method of claim 17, wherein the plurality of consumer records is received in real time.

19. The computer implemented method of claim 17, wherein the plurality of consumer records relate to potential prospective customers that may be approved for credit.

20. The computer implemented method of claim 17, wherein the plurality of consumer records is received in a batch.

21. The computer implemented method of claim 17, wherein the plurality of consumer records represents existing consumer accounts.

22. The computer implemented method of claim 17, wherein the bust out model predicts fraud one to three months in advance.

23. The computer implemented method of claim 17, wherein the credit bureau data is accessible from one or more data sources.

24. A non-transitory storage medium having a computer program stored thereon for causing a suitably programmed system to process computer-program code by performing a method when the computer program is executed on the suitably programmed system, the method comprising:

identifying a plurality of consumer records;

receiving a bust out filter from a storage repository, the bust out filter created based on substantially only credit bureau data;

applying the bust out filter to each of the plurality consumer records to generate a bust out score for each of the plurality of consumer records; and

storing in a data storage the bust out score associated with each of the plurality of consumer records,

wherein the bust out score indicates whether a consumer record exhibits bust out fraud, wherein bust out fraud exhibits a consumer opening one or more credit accounts across one or more credit issuers and initially exhibiting normal consumer behavior followed by a large number of purchases, cash advances, or other uses of credit before abandonment of the one or more accounts.

25. The storage medium of claim 24, wherein the credit bureau data is accessible from one or more data sources.