Risk assessment rule set application for fraud prevention

• US 8,924,279 B2
• Filed: 05/05/2010
• Issued: 12/30/2014
• Est. Priority Date: 05/07/2009
• Status: Active Grant

• Current Assignee: Visa USA, Inc. (Visa, Inc.)
• Original Assignee: Visa USA, Inc. (Visa, Inc.)
• Inventors: Liu, Frederick, Hilgers, Nancy, Nelsen, Mark, Siegel, Kevin, Brown, Christopher, Dwyer, Thomas
• Primary Examiner(s): Maguire, Lindsay M.
• Application Number:
  US12/774,521
• Publication Number:
  US 20100287099A1
• Time in Patent Office: 1,700 Days
• Field of Search: 705/37
• US Class Current: 705/37
• 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 20/405 Establishing or using transaction specific rules

Abstract

Rules, applied to deny authorization of likely fraudulent transactions, are derived from a modified Patient Rule Induction Method algorithm that uses a target variable and a data set of past transactions each associated with a plurality of input variables and a hyper-rectangle enclosing a multi-dimensional space defined by a representation of the input variable values as points within the multi-dimensional space. While a count of the points within the hyper-rectangle is greater than a minimum support parameter, a first plurality of points proximal to edges of the hyper-rectangle are removed, where each such removing maximizes a mean value of the target variable, and then, while the mean value remains maximized, a second plurality of points proximal to the edges is added, where each adding maximizes or maintains the mean value. The hyper-rectangle is bounded within a minimum bounding box that defines the rules.

Claims

1. A computer-implemented method for determining and for denying the authorization of fraudulent financial transactions, the method comprising:

receiving, via a communication network, a request to authorize a transaction between a merchant and an account holder on an account issued by an issuer, wherein the transaction includes associated transaction data;

generating, via one or more processors, a hyper-rectangle enclosing a multi-dimensional space defined by a plurality of input variable values, wherein;

i) the plurality of input variable values respectively correspond to a plurality of transaction parameter input variables that are represented as points within the multi-dimensional space,

ii) the hyper-rectangle includes a plurality of edges, each edge associated with one of the plurality of transaction parameter input variables,

iii) the plurality of transaction parameter input variables is associated with a data set, and

iv) the data set corresponds to information from a plurality of historical transactions, wherein each historical transaction was conducted between one of a plurality of merchants and one of a plurality of account holders on a corresponding account of the one of plurality of account holders issued by one of a plurality of issuers;

retrieving, via the one or more processors, a target optimization variable associated with the data set, wherein the target optimization variable i) is one of the plurality of transaction parameter input variables to be optimized to a specified and desired threshold and ii) represents an output value based on at least one of the plurality of input variable values;

in response to the determination that a number of the points within the hyper-rectangle is greater than a minimum support parameter, removing, via the one or more processors, a first plurality of points proximal to the plurality of edges, wherein said removing maximizes a mean value of the target optimization variable for the hyper-rectangle;

in response to the determination that the mean value of the target optimization variable is maximized, adding, via the one or more processors, a second plurality of points proximal to the plurality of edges, wherein said adding maximizes or maintains the mean value of the target optimization variable;

bounding, via the one or more processors, the hyper-rectangle within a minimum bounding box, the minimum bounding box defining dimensions involved at least in the removal of the first plurality of points and the addition of the second plurality of points;

determining, via the one or more processors, a plurality of fraud prevention boundary limits based on the determined minimum bounding box;

determining, via the one or more processors, whether the transaction is fraudulent based on the application of the plurality of fraud prevention boundary limits on the received transaction data; and

in response to the determination that the transaction is fraudulent, transmitting, via the communication network, a denial response to the request to authorize the transaction.

2. The method as defined in claim 1, further comprising in response to the determination that the transaction is fraudulent, notifying the issuer of the account.

3. The method as defined in claim 1, further comprising:

determining whether the transaction is eligible for backup processing; and

in response to the determination that the transaction is eligible for backup processing, applying a backup processing procedure to process the transaction.

4. The method as defined in claim 3, further comprising:

in response to the determination that if the transaction is not eligible for backup processing, referring for processing the transaction to the issuer of the account upon which the transaction was being conducted.

5. The method as defined in claim 1, further comprising:

identifying, from the transaction data, at least one of i) an account number and ii) Bank Identification Number (BIN); and

determining whether the at least one of the account number and the BIN is a real-time decisioning participant;

in response to the determination that the at least one of the account number and the BIN is a real-time descisioning participant, determining whether the transaction is fraudulent in real-time.

6. The method as defined in claim 1, wherein the fraud prevention boundary limits are card-level real-time decisioning rules, the method further comprising:

obtaining access to a card database;

identifying an account number from the transaction data; and

determining whether the account number is stored within the card database, wherein when the account number is stored within the card database the plurality of fraud prevention boundary limits are applied to the transaction data.

7. The method as defined in claim 1, wherein the fraud prevention boundary limits are Bank Identification Numbers (BIN) level rules, the method further comprising:

obtaining access to a BIN database;

identifying a BIN from the transaction data; and

determining whether the BIN is stored within the BIN database, wherein when the BIN is stored within the BIN database the plurality of fraud prevention boundary limits are applied to the transaction data.

8. The method as defined in claim 1, further comprising obtaining the minimum support parameter.

9. The method as defined in claim 1, further comprising obtaining a maximum peeling parameter, wherein the maximum peeling parameter is the maximum count of the first plurality of points that is removable during said removing.

10. The method as defined in claim 1, further comprising obtaining a condition associated with at least one of the plurality of transaction parameter input variables, wherein the multi-dimensional space is defined by the plurality of input variable values that satisfy the condition.

11. The method as defined in claim 1, wherein the plurality of fraud prevention boundary limits correspond to the edges of the minimum bounding box.

12. The method as defined in claim 1, wherein the transaction parameter input variable to be optimized is one of:

a mean sale amount for a merchant, a number of chargebacks;

a transaction volume; and

a number of fraudulent transactions.