Mortgage fraud detection systems and methods

• US 7,546,271 B1
• Filed: 07/16/2008
• Issued: 06/09/2009
• Est. Priority Date: 12/20/2007
• Status: Expired due to Fees

• Current Assignee: LexisNexis Risk Solutions, Inc. (RELX PLC)
• Original Assignee: Choicepoint Asset Company
• Inventors: James, Denise, Chmielewski, Thomas
• Primary Examiner(s): Trammell; James P
• Assistant Examiner(s): Swartz; Jamie H
• Application Number:
  US12/174,591
• Publication Number:
  US 20090164232A1
• Time in Patent Office: 328 Days
• Field of Search: 705/1, 705/35, 705/38, 705/40, 705/4
• US Class Current: 705/38
• CPC Class Codes: G06Q 30/0185 Product, service or business identity fraud
• 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: G06Q 50/16 Real estate
• CPC Class Codes: G06Q 50/18 Legal services

Abstract

Mortgage fraud detection systems and methods are provided. A fraud detection system can generally comprise a database operator, a database manager, one or more users, and a database. Database components can be implemented as hardware, software, or a combination of both. A fraud detection method can generally comprise providing a database for maintaining a plurality of records. Data records can be continually received by a database in batchwise or single submission protocols. Records can be compared to determine whether they contain common data in one or more predetermined fields. If records contain common data in one or more predetermined fields, records can be examined for inconsistencies in one or more other data fields. Discovered inconsistencies and magnitudes of inconsistencies can be warning signs of fraud. Results of comparisons can be reported to system users to enable such users to investigate additional circumstances. Other embodiments and features are also claimed and described in the application.

Claims

1. A fraud detection system to detect fraud in a financial loan application process, the system comprising:

one or more interfaces configured to receive loan application information data from a plurality of unique data sources that comprise a plurality of different lenders, the loan application information data comprising a first plurality of loan application data records originating from a first lender and a second plurality of loan application data records originating from a second lender, wherein the first plurality of loan application data records comprises at least a first loan application data record originating from the first lender;

a database configured to receive the loan application information data from the plurality of unique data sources and to store the loan application information data with an identifier configured to distinguish loan application information data records received from the plurality of unique data sources;

a computer processor configured to populate an internal pipeline for the first lender comprising the first plurality of loan application data records originating from the first lender by using the unique identifier associated with the first lender;

the computer processor further configured to populate an external pipeline for the first lender comprising at least the second plurality of loan application data records originating from the second lender;

the computer processor further configured to compare the first loan application data record to other loan application data records in the first lender'"'"'s internal pipeline, and further configured to compare the first loan application data record to loan application data records in the first lender'"'"'s external pipeline; and

the computer processor further configured to identify potential fraud in a loan application corresponding to the first loan application data record based on the comparisons between the first loan application data record and the first lender'"'"'s internal and external pipelines.

2. The fraud detection system of claim 1, the computer processor further configured to determine whether one or more loan application data records contain one or more duplicate data fields as compared to the first data record, and to flag the loan application data records containing the one or more duplicate data fields;

and

the computer processor further configured to determine if one or more of the flagged loan application data records contain a discrepancy in one or more other data fields as compared to the first loan application.

3. The fraud detection system of claim 2, the computer processor further configured to report at least one of information concerning the loan application information data records containing the one or more duplicate data fields and information concerning the flagged loan application information data records containing discrepant information in the one or more other data fields.

4. The fraud detection system of claim 3, the computer processor further configured to organize information in a report at least partially based on a weight corresponding to the one or more other data fields and a measurement of discrepancy in the one or more other data fields.

5. The fraud detection system of claim 1, the computer processor further configured to place one or more received loan application information data records in an active state for a predetermined holding period and periodically analyze active-state loan application data records relative to loan application data records received after the active-state data records.

6. The fraud detection system of claim 1, wherein the computer processor is configured to determine if one or more of the loan application data records contain substantially similar addresses for a real estate property involved in a loan transaction.

7. The fraud detection system of claim 1, the computer processor further configured to conduct an internal pipeline loan analysis by comparing loan application data records from the first lender to other loan one another and an external loan pipeline analysis by comparing loan application data records from other lenders to one another.

8. The fraud detection system of claim 1, the database further configured to receive the loan application information data at an initial loan application stage of a loan application process.

9. A computer-implemented method to detect potential fraud in a financial credit or loan application process, the fraud detection method comprising:

storing in a database data comprising a plurality of financial credit or loan applications from a plurality of lenders;

performing a two-level search through use of a computer processor, the two-level search comprising;

identifying a subset of loan applications containing two or more of the loan applications having duplicate data in one or more predetermined data fields; and

determining whether the loan applications in the subset of loan applications comprise one or more other fields having discrepant data;

identifying potential instances of fraud based on the discrepant data; and

conducting a dual loan pipeline analysis that includes an internal pipeline analysis comprising a comparison of all loan applications from a first lender to one another and an external pipeline analysis comprising a comparison of loan applications from the first lender to loan applications from the plurality of lenders.

10. The fraud detection method of claim 9, wherein identifying a subset of loan applications containing one or more of the loan applications having duplicate data in one or more predetermined data fields comprises determining if the one or more data records contain data fields having duplicate property address information or duplicate borrower information.

11. The fraud detection method of claim 9, further comprising providing reports to a user with data organized based on results of a sorting algorithm providing weights to data fields of the data.

12. The fraud detection method of claim 9, further comprising holding data pertaining to one or more of the loan applications in the database for a predetermined holding period and, with a computer processor, comparing existing loan applications to loan applications received after the existing loan applications at predefined intervals.

13. The fraud detection method of claim 9, further comprising performing the two-level search on the loan applications during an initial loan application stage.

14. The fraud detection method of claim 9, further comprising identifying duplicate data between a first data field of a first loan application and a second data field of a second loan application, wherein the first data field and the second data field represent different elements of a loan application.

15. The fraud detection method of claim 9, the method further comprising determining a type of fraud for the first loan application through an iterative process comprising, for each fraud type of a predetermined plurality of fraud types, determining how closely one or more characteristics of the first loan application meet predetermined criteria corresponding to the fraud type.

16. A computer-readable medium having computer readable instructions stored thereon for execution by a processor to execute a method for identifying potential fraud in a loan application, the method comprising:

receiving data related to a first loan application originating from a first subscriber;

receiving data related to a second loan application, the first loan application and the second loan application having a duplicate value in a first data field, and having disparate values in a second data field, wherein the first data field corresponds to one of a plurality of predetermined key data fields;

configuring a computer processor to compare the first loan application to the second loan application;

flagging the second loan application based on identification of the duplicate value in the first data field and the disparate values in the second data field;

inserting the second loan application into a set of flagged loan applications, wherein each of the set of flagged loan applications comprises a duplicate value in a key data field and a disparate value in an other data field, as compared to the first loan application;

ranking the first loan application among a plurality of loan applications based on a predetermined algorithm, wherein the first loan application is ranked relative to the plurality of loan applications based at least partially on (i) importance of the other data fields having disparate values in the set of flagged loan applications and (ii) a measurement of disparity between the disparate values in the other data fields of the set of flagged loan applications, as compared to the first loan application; and

reporting the ranked plurality of loan applications to the first subscriber.

17. The computer-readable medium of claim 16, the method further comprising calculating a score for the first loan application, wherein the score is based on results of comparing the first loan application to other loan applications in the set of flagged loan applications.

18. The computer-readable medium of claim 16, wherein the first data field containing the duplicate value comprises property address information, and wherein the first data field is examined to determine if a single property is subject to multiple loan applications.

19. The computer-readable medium of claim 16, the method further comprising repeatedly determining on a pre-determined basis whether the first loan application has a data field containing a duplicate value of another data field contained within at least one other received loan application, and whether the received loan applications having the duplicate values as compared to the first loan application have other data fields with disparate data fields as compared to the first loan application.