Method and apparatus for pattern generation

• US 6,018,723 A
• Filed: 05/27/1997
• Issued: 01/25/2000
• Est. Priority Date: 05/27/1997
• Status: Expired due to Term

• Current Assignee: Visa International (Visa, Inc.)
• Original Assignee: Visa International Service Association (Visa, Inc.)
• Inventors: Faith, Patrick L., Siegel, Kevin P., Washburne, Theodore P.
• Primary Examiner(s): Trammell, James P.
• Assistant Examiner(s): ROSEN, NICHOLAS D
• Application Number:
  US08/863,443
• Time in Patent Office: 973 Days
• Field of Search: 364/408, 705/38, 705/39, 705/42, 705/43, 705/44, 705/1
• US Class Current: 705/38
• CPC Class Codes: G06Q 20/10 specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
• CPC Class Codes: G06Q 20/108 Remote banking, e.g. home banking
• 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 20/403 Solvency checks
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/08 Insurance
• 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

Methods and apparatus for transforming scoreable transaction data into financial data features are disclosed. In one aspect, a computer-implemented method transforms transaction data into a financial data feature for assessing credit risks. The financial data feature is extracted from the transaction data. The method involves obtaining the transaction data from a data source, and performing a set of operations on the transaction data to transform the transaction data into the financial data feature. The set of operations is selected only from a predefined set of classes of operations which are interrelated by a predefined order of precedence. Each operation in the set of operations is performed in an order based on the predefined order of precedence of a class associated with each operator.

Claims

1. A computer-implemented method for transforming scoreable transaction data into a financial data feature for use in assessing credit risk, the financial data feature being extracted from the scoreable transaction data, the method comprising:

obtaining the scoreable transaction data from a data source; and

performing a set of operations on the scoreable transaction data to transform the scoreable transaction data into the financial data feature, the set of operations being selected only from a predefined set of classes of operations, the set of predefined classes of operations being arranged in a predefined order of precedence, wherein each operation in the set of operations is performed in an order based on the predefined order of precedence of a class associated with the each operator, the set of predefined classes of operations including at most five classes of operations, the five classes of operations being a data structure class, an atomic transformation class, an entity transformation class, a time transformation class, and a joining operator class, wherein performing the set of operations includes first performing a data structure operation associated with the data structure class.

2. The computer-implemented method of claim 1 further including:

performing an atomic transformation operation associated with the atomic transformation class directly after performing the data structure operation associated with the data structure class;

performing an entity transformation operation associated with the entity transformation class directly after performing the atomic transformation operation associated with the atomic transformation class;

performing a time transformation operation associated with the time transformation class directly after performing the entity transformation operation associated with the entity transformation class; and

performing joining operator operation associated with the joining operator class directly after performing the entity transformation operation associated with the entity transformation class.

3. A computer-implemented method for implementing a pattern generator, the pattern generator being associated with a predictive process, the pattern generator being arranged to transform a given transaction, the method comprising:

obtaining a data stream, the data stream including at least one data field containing information associated with the given transaction; and

performing at least one overall operation on the information to transform the information into a characteristic variable, the overall operation being selected from a set of five classes associated with the pattern generator, wherein when there is more than one overall operation to be performed, a first overall operation and a second overall operation are performed based on an order of precedence associated with the five classes, whereby the characteristic variable is arranged to be accepted as an input into the predictive process.

4. A computer-implemented method as recited in claim 3 wherein the set of five classes includes:

a first class which includes at least one operation arranged to relate the information to a database;

a second class which includes at least one operation arranged to transform the information from a first format into an atomic format;

a third class which includes at least one operation arranged to transform the information from an atomic format into a second format;

a fourth class which includes at least one operation arranged to perform calculations using the information; and

a fifth class which includes at least one operation arranged to relate the information to additional information.

5. A computer-implemented method as recited in claim 4 further including performing a first operation in the first class.

6. A computer-implemented method as recited in claim 5 further including:

performing a first operation in the second class directly after performing the first operation in the first class; and

performing one of a second operation in the first class, a first operation in the third class, a first operation in the fourth class, and a first operation in the fifth class, directly after performing the first operation in the second class, as specified by the order of precedence.

7. A computer-implemented method as recited in claim 5 further including:

performing a first operation in the third class directly after performing the first operation in the first class; and

performing one of a second operation in the first class, a second operation in the third class, a first operation in the fourth class, and first operation in the fifth class, directly after performing the first operation in the third class, as specified by the order of precedence.

8. A computer-implemented method as recited in claim 5 further including:

performing a first operation in the fourth class directly after performing the first operation in the first class; and

performing one of a second operation in the first class, a second operation in the fourth class, a first operation in the third class, and first operation in the fifth class, directly after performing the first operation in the fourth class, as specified by the order of precedence.

9. A computer-implemented method as recited in claim 5 further including:

performing a first operation in the fifth class directly after performing the first operation in the first class; and

performing one of a second operation in the first class and a first operation in the fourth class directly after performing the first operation in the fifth class, as specified by the order of precedence.

10. A pattern generation engine arranged to transform a given transaction into a characteristic variable, the given transaction being associated with a data stream including at least one data field, the characteristic variable being associated with a modeling process, the pattern generation engine including:

a first class which includes at least one operation arranged to relate the data field to a database;

a second class which includes at least one operation arranged to transform the data field from a first format into an atomic format;

a third class which includes at least one operation arranged to transform the data field from an atomic format into a second format;

a fourth class which includes at least one operation arranged to perform calculations using the data field; and

a fifth class which includes at least one operation arranged to relate the data field to additional information, wherein the characteristic variable is generated using at least one of the at least one operation in the first class, the at least one operation in the second class, the at least one operation in the third class, the at least one operation in the fourth class, and the at least one operation in the fifth class.

11. A pattern generation engine as recited in claim 10 wherein the first class, the second class, the third class, the fourth class, and the fifth class are related by a precedence relationship.

12. A pattern generation engine as recited in claim 11 wherein:

the precedence relationship is arranged to enable the at least one operation in the first class to precede the at least one operation in the second class;

the precedence relationship is arranged to enable the at least one operation in the first class to precede the at least one operation in the third class;

the precedence relationship is arranged to enable the at least one operation in the first class to precede the at least one operation in the fourth class; and

the precedence relationship is arranged to enable the at least one operation in the first class to precede the at least one operation in the fifth class.

13. A pattern generation engine as recited in claim 12 wherein:

the precedence relationship is arranged to enable a first operation in the third class to precede a second operation in the third class; and

the precedence relationship is arranged to enable a first operation in the fourth class to precede a second operation in the fourth class.

14. A pattern generation engine as recited in claim 12 wherein:

the precedence relationship is arranged to enable a first operation in the second class to precede a first operation in the first class;

the precedence relationship is arranged to enable a second operation in the second class to precede a first operation in the third class;

the precedence relationship is arranged to enable a third operation in the second class to precede a first operation in the fourth class; and

the precedence relationship is arranged to enable a fourth operation in the second class to precede a first operation in the fifth class.

15. A pattern generation engine as recited in claim 12 wherein:

the precedence relationship is arranged to enable a first operation in the third class to precede a first operation in the fifth class; and

the precedence relationship is arranged to enable a second operation in the fifth class to precede a second operation in the third class.

16. A pattern generation engine as recited in claim 12 wherein:

the precedence relationship is arranged to enable a first operation in the third class to precede a first operation in the fourth class; and

the precedence relationship is arranged to enable a second operation in the fourth class to precede a second operation in the third class.

17. A pattern generation engine as recited in claim 12 wherein:

the precedence relationship is arranged to enable a first operation in the fourth class to precede a first operation in the fifth class; and

the order of precedence enables a second operation in the fifth class to precede a second operation in the fourth class.

18. A pattern generation engine as recited in claim 10 wherein:

the given transaction is a financial transaction; and

the modeling process is a predictive modeling process.

19. A pattern generation engine as recited in claim 18 wherein the modeling process is a risk predictive modeling process.

20. A pattern generation engine as recited in claim 19 wherein the given transaction is associated with a first entity, and the modeling process is arranged to assess a risk of bankruptcy for the first entity based on the given transaction.

21. A pattern generation engine as recited in claim 10 wherein:

the given transaction is a financial transaction; and

the modeling process is a transaction scoring process.