System and method for detecting fraudulent transactions

• US 20030069820A1
• Filed: 11/18/2002
• Published: 04/10/2003
• Est. Priority Date: 03/24/2000
• Status: Active Grant

• Current Assignee: Access Business Group International LLC (Alticor Incorporated)
• Original Assignee: Amway Corporation (Alticor Incorporated)
• Inventors: Jones, Richard A., Deshpande, Sanjay H., Lewis, Kelly R. Jr., Gessner, Colleen M., Hillmer, James M., Johnston, Connie S.
• Granted Patent: US 6,714,918 B2
• Time in Patent Office: Days
• US Class Current: 705/35
• CPC Class Codes: G06Q 20/24 Credit schemes, i.e. "pay after"
• CPC Class Codes: G06Q 20/40 Authorisation, e.g. identification of payer or payee, verification of customer or shop credentials; Review and approval of payers, e.g. check credit lines or negative lists
• CPC Class Codes: G06Q 20/4016 involving fraud or risk level assessment in transaction processing
• CPC Class Codes: G06Q 20/4037 Remote solvency checks
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• 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: G07F 7/08 by coded identity card or credit card or other personal identification means without personal verification means G07F7/02

Abstract

A system for detecting fraudulent transactions is disclosed. The system breaks the transaction down into its component parameters. A first score is computed depending on the propensity of the transacted commodity to be involved in fraud. A second score is computed as a function of the authentication of the remaining parameters of the transaction. A total fraud score is computed from the first score and the second score and compared against a fraud threshold to determine the likelihood of the transaction being fraudulent.

Claims

1. A method for detecting fraudulent transactions between a customer and a vendor comprising:

(a) receiving a plurality of transaction parameters from said vendor, said plurality of transaction parameters representing at least one transaction of at least one commodity between said customer and said vendor;

(b) computing a first score as a function of said at least one commodity;

(c) computing a second score as a function of a first one or more of said plurality of transaction parameters;

(d) computing a fraud score based on said first and second scores; and

(e) indicating to said vendor that said at least one transaction is fraudulent if said fraud score exceeds a first pre-determined threshold.

2. The method of claim 1 further comprising:

(f) comparing a second one or more of said plurality of transaction parameters to a positive database; and

(g) indicating that said at least one transaction is non-fraudulent if a match results, said positive database comprising past non-fraudulent transactions known to said vendor.

3. The method of claim 2, wherein a subset of (a)-(g) are performed by said vendor and the remaining are performed by an entity external to said vendor.

4. The method of claim 2, further comprising:

(h) adding one or more of said plurality of transaction parameters to said positive database if said comparing fails to match and said fraud score does not exceed said first pre-determined threshold.

5. The method of claim 1 further comprising:

(f) comparing a second one or more of said plurality of transaction parameters to a negative database; and

(g) indicating that said at least one transaction is fraudulent if a match results, said negative database comprising known past fraudulent transactions.

6. The method of claim 5, wherein a subset of (a)-(g) are performed by said vendor and the remaining are performed by an entity external to said vendor.

7. The method of claim 5, further comprising:

(h) adding one or more of said plurality of transaction parameters to said negative database if said fraud score exceeds said first pre-determined threshold.

8. The method of claim 1 further comprising:

(f) computing a total value of all of said at least one commodity; and

(g) indicating that said at least one transaction is non-fraudulent if said total value is less than a second predetermined threshold.

9. The method of claim 8, wherein said second pre-determined threshold is approximately $100.

10. The method of claim 8, wherein (g) further comprises dynamically adjusting said second pre-determined threshold.

11. The method of claim 8, wherein a subset of (a)-(g) are performed by said vendor and the remaining are performed by an entity external to said vendor.

12. The method of claim 8, wherein (c) further comprises adjusting said second score as a function of said total value.

13. The method of claim 8, wherein (e) further comprises adjusting said first pre-determined threshold as a function of said total value.

14. The method of claim 1 wherein a subset of (a)-(e) are performed by said vendor and the remaining are performed by an entity external to said vendor.

15. The method of claim 1, wherein said at least one commodity is characterized by a likelihood of fraud and further wherein said first score is a function of said likelihood.

16. The method of claim 1, wherein said plurality of transaction parameters are received after said at least one transaction has been completed.

17. The method of claim 1, wherein (a)-(e) are completed while said at least one transaction is pending.

18. The method of claim 1, wherein (b) further comprises referencing a database, said database including a plurality of commodities and corresponding scores.

19. The method of claim 1, wherein (c) further comprises a summation function.

20. The method of claim 1, wherein (d) further comprises referencing a database, said database comprising data corresponding to a plurality of statistical data, a plurality of consumer data and a plurality of historical transaction data, and further wherein said second score is computed as a function of a pre-determined point value for each of said plurality of transaction parameters and whether said first one or more of said plurality of transaction parameters is authenticated by said data.

21. The method of claim 20, wherein said pre-determined point value is less than zero.

22. The method of claim 20, further comprising:

(f) adjusting said second score based on said pre-determined point values of a subset of said first one or more of said plurality of transaction parameters.

23. The method of claim 1, wherein (e) further comprises adjusting said first pre-determined threshold as a function of said second score.

24. The method of claim 1, wherein (d) further comprises a summation function.

25. The method of claim 1, wherein (e) further comprises multiplying said first score by said second score.

26. The method of claim 1, wherein said vendor is a catalog based merchant.

27. The method of claim 1, wherein said vendor is an electronic commerce based merchant.

28. The method of claim 1, wherein said vendor is a store based retail merchant.

29. A fraud detection processor comprising:

a receiver operative to receive a plurality of transaction parameters from a vendor, said plurality of transaction parameters representing at least one transaction of one or more commodities between a customer and said vendor;

a first score processor coupled with said receiver and operative to compute a first score as a function of each of said one or more commodities;

a second score processor coupled with said receiver and operative to compute a second score as a function of each of a first one or more of said plurality of transaction parameters;

a fraud score processor coupled with said first and second score processors and operative to compute a fraud score as a function of said first and second scores;

fraud determination logic coupled with said fraud score processor and operative to compute a determination of whether said at least one transaction is fraudulent as a function of a comparison of said fraud score and a first pre-determined threshold and further operative to indicate said determination to said vendor.

30. The fraud detection processor of claim 29 further comprising:

a positive database comprising past non-fraudulent transactions known to said vendor; and

a parameter processor coupled with said receiver and said fraud determination logic and operative to compare a second one or more of said plurality of transaction parameters to said positive database and indicate a result of said comparison to said fraud determination logic;

wherein said fraud determination logic is further operative to determine that said at least one transaction is not fraudulent when said parameter processor indicates a match result.

31. The fraud detection processor of claim 30, wherein said parameter processor is further operative to add one or more said plurality of transaction parameters to said positive database if said comparison fails to match and said fraud determination logic determines that said at least one transaction is non-fraudulent.

32. The fraud detection processor of claim 29 further comprising:

a negative database comprising data related to known past fraudulent transactions; and

a parameter processor coupled with said receiver and said fraud determination logic and operative to compare a second one or more of said plurality of transaction parameters to said negative database and indicate a result of said comparison to said fraud determination logic;

wherein said fraud determination logic is further operative to determine that said at least one transaction is fraudulent when said parameter processor indicates a match result.

33. The fraud detection processor of claim 32, wherein said parameter processor is further operative to add one or more of said plurality of transaction parameters to said negative database if said comparison fails to match and said fraud determination logic determines that said at least one transaction is fraudulent.

34. The fraud detection processor of claim 29, further comprising

a valuation processor coupled with said receiver and fraud determination logic and operative to compute a total value of all of said at least one commodity and indicate to said fraud determination logic if said total value is less than a second pre-determined threshold;

wherein said fraud determination logic is further operative to determine that said at least one transaction is non-fraudulent when said valuation processor indicates that said total value is less than said second pre-determined threshold.

35. The fraud detection processor of claim 34, wherein said second pre-determined threshold is approximately $100.

36. The fraud detection processor of claim 34, wherein said second pre-determined threshold is dynamic.

37. The fraud detection processor of claim 34, wherein said valuation processor is further coupled with said second score processor and further wherein said second score processor is further operative to adjust said second score based upon said total value.

38. The fraud detection processor of claim 34, wherein said fraud determination logic is further operative to adjust said first pre-determined threshold as a function of said total value.

39. The fraud detection processor of claim 29, wherein said at least one commodity is characterized by a probability of fraud and further wherein said first score is as a function of said probability.

40. The fraud detection processor of claim 29, wherein said plurality of transaction parameters are received after said at least one transaction has been completed.

41. The fraud detection processor of claim 29, wherein said fraud determination logic is further operative to make said determination while said at least one transaction is pending.

42. The fraud detection processor of claim 29, wherein said first score processor further comprises a commodity reference database, said commodity reference database including a plurality of commodities and corresponding scores wherein said scores are a function of the probability of fraudulent activity related to a corresponding commodity.

43. The fraud detection processor of claim 29, wherein said second score processor further comprises a transactional reference database, said transactional reference database comprising data corresponding to a plurality of statistical data, a plurality of consumer data and a plurality of historical transaction data, and further wherein said second score processor is further operative to compute said second score as a function of a pre-determined point value for each of said plurality of transaction parameters and whether said first one or more of said plurality of transaction parameters is authenticated by said data in said transactional reference database.

44. The fraud detection processor of claim 43, wherein said pre-determined point value is less than zero.

45. The fraud detection processor of claim 43, wherein said second score processor is further operative to adjust said second score as a function of said pre-determined point values of a sub-set of said plurality of transaction parameters.

46. The fraud detection processor of claim 29, wherein said fraud determination logic is further operative to adjust said first pre-determined threshold based on said second score.

47. The fraud detection processor of claim 29, wherein said vendor is a catalog based merchant.

48. The fraud detection processor of claim 29, wherein said vendor is an electronic commerce based merchant.

49. The fraud detection processor of claim 29, wherein said vendor is a store based retail merchant.

50. A computer implemented system for detecting fraudulent transactions between a customer and a vendor comprising:

a positive database comprising data representing past non-fraudulent transactions known to said vendor;

a negative database comprising data representing past known fraudulent transactions;

one or more transaction databases comprising consumer and geographic transaction information;

a commodity database comprising data representing valuation and propensity to incite fraudulent activity for a plurality of commodities;

transaction logic operative to receive a transaction of a commodity from a vendor and decompose said transaction into a plurality of transaction parameters;

result logic operative to send an indication to said vendor whether said transaction is fraudulent and non-fraudulent;

valuation logic coupled with said transaction logic, said result logic and said commodity database and operative to compute the total value of said transaction and cause said result logic to indicate that said transaction is non-fraudulent if said total value is less than a first pre-determined threshold;

comparison logic coupled with said transaction logic, said result logic and said positive database and operative to compare said transaction with said positive database and cause said result logic to indicate that said transaction is non-fraudulent if said transaction matches in said positive database; and

fraud logic coupled with said transaction logic, said result logic and said negative database, said one or more transaction databases and said commodity database and operative to compute a first score as a function of said past fraudulent history and said propensity to incite fraudulent activity of said commodity, a second score as a function of the authenticity of said plurality of transaction parameters and a fraud score as a function of said first and second scores, said fraud logic being further operative to cause said result logic to indicate that said transaction is fraudulent if said fraud score exceeds a second pre-determined threshold.

51. The system of claim 50, wherein one or more of said positive database, said negative database, said one or more transaction databases, said commodity database, said transaction logic, said result logic, said valuation logic, said comparison logic and said fraud logic operates on a computer system internal to said vendor and the remaining operates on a computer system external to said vendor, said internal and external computer systems in communication with each other.

52. The system of claim 50, wherein said fraud logic is further coupled with said valuation logic and further wherein said second score is dynamically adjusted as a function of said total value.

53. The system of claim 50, wherein said fraud logic is further coupled with said valuation logic and further wherein said second predetermined threshold is dynamically adjusted as a function of said total value.

54. The system of claim 50, wherein said system is operative to process said transaction in real time.

55. The system of claim 50, wherein said system is operative to batch process said transaction.