Method and apparatus for detecting fraud

• US 6,253,186 B1
• Filed: 08/14/1996
• Issued: 06/26/2001
• Est. Priority Date: 08/14/1996
• Status: Expired due to Term

• Current Assignee: Blue Cross And Blue Shield Of South Carolina
• Original Assignee: Blue Cross And Blue Shield Of South Carolina
• Inventors: Pendleton, E. Steele Jr.
• Primary Examiner(s): Millin, Vincent
• Assistant Examiner(s): Kanof, Pedro R.
• Application Number:
  US08/696,615
• Time in Patent Office: 1,777 Days
• Field of Search: 705/1, 705/2, 705/3, 705/7, 705/10, 705/11
• US Class Current: 705/2
• CPC Class Codes: G06Q 10/06393 Score-carding, benchmarking or key performance indicator [KPI] analysis
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/0185 Product, service or business identity fraud
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

A computerized arrangement for detecting potentially fraudulent suppliers or providers of goods or services includes a processor, a storage device, an input device for communicating data to the processor and storage device, and an output device for communicating data from the processor and storage device. The storage device includes a claims data file for storing information relating to a plurality of claims submitted for payment by a selected supplier or provider, one or more encoding lookup tables for use with the claims data file to produce an encoded claims data file, and a neural network program for analyzing the encoded data to produce an indicator of potentially fraudulent activity. The indicator may be compared to a predetermined threshold value by the apparatus or method to identify fraudulent suppliers. In addition to the neural network, at least one expert system may be used in the identification process.

Claims

1. A computerized method for analyzing the billing patterns of a plurality of providers and suppliers of goods and services, and for identifying potentially fraudulent providers and suppliers, comprising the steps of:

a. training a neural network to recognize patterns associated with fraudulent billing activity by providers and suppliers of goods and services;

b. collecting data from said plurality of providers and suppliers, including claims data relating to claims submitted for payment by said providers and suppliers;

c. extracting adjudicated line data from said claims data to construct a claim file;

d. analyzing the line data in the claim file by means of the neural network to produce a numerical fraud indicator for at least one of said providers and suppliers;

e. comparing the fraud indicator to a predetermined threshold indicator value;

f. storing an identifier associated with each provider or supplier whose fraud indicator exceeds the predetermined threshold indicator value;

g. analyzing data relating to each of the providers or suppliers whose identifiers were stored in step f, using an expert system inference engine in accordance with a plurality of expert system rules, to identify potentially fraudulent providers or suppliers; and

h. producing a report which displays the identities of the potentially fraudulent providers or suppliers.

2. The method of claim 1, comprising the additional step of using the report to initiate further investigation of the potentially fraudulent providers or suppliers identified in step g.

3. The method of claim 1, further comprising the additional steps of examining previously stored statistical information relating to each provider or supplier whose fraud indicator exceeds the predetermined threshold indicator value, and storing results of the examination in a utilization screening file.

4. The method of claim 3, comprising the additional step of preparing a report displaying the results of the examination of previously stored statistical information.

5. The method of claim 1, wherein the data analyzed in step g includes data from at least one of a physical traits analysis file, a statistical screening file, and a utilization screening file.

6. The method of claim 1, comprising the additional steps of periodically updating and training the neural network, and periodically refining the expert system rules.

7. The method of claim 1, further comprising the step of determining whether the neural network has been trained to analyze a particular item of line data by examining an HCPCS code associated with said data.

8. The method of claim 1, further comprising the step of determining whether or not a decision has been made to pay an adjudicated line from the claim file, and analyzing (step d) only those lines for which a decision to pay has been made.

9. The method of claim 1, wherein the step of analyzing the line data (step d) further includes the steps of:

d(1). selecting elements of information from the data stored in the claim file;

d(2). encoding the selected elements of information to produce an encoded claim file; and

d(3). storing the encoded claim file.

10. The method of claim 9, further comprising the step of sorting the encoded claim file by supplier or provider to produce a sorted encoded claim file.

11. The method of claim 10, wherein the step of analyzing the line data (step d) further comprises the steps of:

d(4). reading data from the sorted encoded claim file; and

d(5). analyzing the data from the sorted encoded claim file by means of the neural network to produce the fraud indicator for the selected supplier or provider.

12. The method of claim 11, wherein the step of analyzing data (step d(5)) includes the steps of:

d(5)(a). producing a plurality of fraud indicators based on a plurality of claims submitted by the selected supplier or provider; and

d(5)(b). computing a composite fraud indicator for the selected supplier or provider.

13. The method of claim 12, wherein the step of computing a composite fraud indicator includes averaging the plurality of fraud indicators for the selected provider or supplier.

14. The method of claim 13, wherein the comparing step includes the step of comparing the composite fraud indicator to the predetermined threshold indicator value.

15. The method of claim 1, comprising the additional step of storing data relating to the identified potentially fraudulent suppliers or providers in at least one of a database file and a statistics file.

16. The method of claim 1, wherein the step of analyzing data (step g) relating to each of the suppliers or providers whose identities were stored in step f comprises at least one of the additional steps of:

g(1). analyzing previously stored statistical information relating to the subject supplier or provider;

g(2). analyzing, by use of a neural network, previously stored physical characteristics relating to the subject provider or supplier; and

g(3). analyzing statistical utilization data relating to the subject supplier or provider.

17. The method of claim 16, wherein at least one of the analysis of previously stored statistical information, the analysis of physical characteristics, and the analysis of statistical utilization data include an analysis by means of an expert system.

18. The method of claim 17, further comprising the additional step of periodically refining a set of rules associated with the expert system in response to data relating to fraudulent suppliers or providers detected by the computerized method.

19. The method of claim 16, further comprising the additional step of periodically updating the neural network used to perform the analysis of supplier or provider physical characteristics.

20. The method of claim 1, comprising the additional step of preparing a report which displays at least one of:

a. the total number of claim lines examined;

b. the total providers or suppliers analyzed;

c. the number of potentially fraudulent providers or suppliers identified; and

d. the number of claim lines examined to produce the fraud indicators for those providers or suppliers whose fraud indicators exceeded the predetermined threshold indicator.

21. A computer system for analyzing the billing patterns of a plurality of providers and suppliers of goods and services, and for identifying potentially fraudulent providers and suppliers, comprising:

a processor;

a storage device for storing a claims data file, and a neural network program trained to recognize patterns associated with fraudulent billing activity by providers and suppliers of goods and services;

input means for receiving data from said plurality of providers and suppliers, including claims data relating to claims submitted for payment by said providers and suppliers, and for communicating said data to the processor and storage device;

means for extracting adjudicated line data from said claims data and for constructing said claims data file from said line data;

means for analyzing the line data in the claims data file by means of the neural network to produce a numerical fraud indicator for at least one of said providers and suppliers;

means for comparing the fraud indicator to a predetermined threshold indicator value;

means for storing an identifier associated with each provider or supplier whose fraud indicator exceeds the predetermined threshold indicator value;

means for analyzing data relating to each of the providers or suppliers whose identifiers were stored to identify potentially fraudulent providers or suppliers, said means comprising an expert system inference engine programmed in accordance with a plurality of expert system rules; and

means for producing a summary report which displays the identities of the potentially fraudulent providers or suppliers.

22. A computer system according to claim 21, wherein said storage device further comprises a neural network database file for storing data relating to the potentially fraudulent suppliers or providers.

23. A computer system according to claim 22, wherein said storage device further comprises a statistics file for storing statistical data relating to a supplier or provider.

24. A computer system according to claim 23, wherein said storage device further comprises a statistical screening file containing data from the neural network database file and the statistics file.

25. A computer system according to claim 22, wherein said storage device further comprises a supplier/provider application file for storing data relating to the physical characteristics of a supplier or provider.

26. A computer system according to claim 25, wherein said storage device further comprises a physical traits analysis file containing data from the neural network database file and the supplier/provider application file.

27. A computer system according to claim 25, wherein said storage device further comprises encoding lookup tables for use with the supplier/provider application file to produce an encoded supplier/provider application data file.

28. A computer system according to claim 27, further comprising means for analyzing the data in the encoded supplier/provider application data file to produce a physical traits analysis file.

29. A computer system according to claim 22, wherein said storage device further comprises a supplier/provider application database for storing information relating to supplier/provider utilization.

30. A computer system according to claim 29, wherein said storage device further comprises a utilization screening file containing data from the neural network database file and the supplier/provider application database.

31. A computer system according to claim 21, wherein said means for analyzing includes means for conducting a fuzzy logic analysis to produce a fraud prediction relating to the identified supplier or provider.

32. A computer system according to claim 21, further comprising means for training the neural network program.

33. Apparatus for detecting potentially fraudulent suppliers or providers of goods or services, comprising:

a processor;

a storage device;

input means for communicating data from an input device to the processor and storage device; and

output means for communicating data from the processor storage device to an output device;

said storage device further comprising;

a neural network database file for storing data relating to potentially fraudulent suppliers or providers;

at least one of;

(I) a physical traits analysis file for storing application data relating to a plurality of suppliers and providers;

(ii) a statistical screening file for storing statistical data relating to potentially fraudulent suppliers or providers; and

(iii) a utilization screening file for storing historical data relating to a plurality of suppliers or providers;

means for processing information from the neural network database file and at least one of the physical traits analysis file, the statistical screening file and the utilization screening file to identify potentially fraudulent suppliers or providers; and

means for producing a summary report which displays the identities of the potentially fraudulent suppliers or providers.

34. Apparatus according to claim 33, wherein said means for processing comprises an expert system, and wherein said storage device further contains a plurality of expert system rules.

35. Apparatus according to claim 33, further comprising means for updating at least one of the physical traits analysis file, the statistical screening file and the utilization screening file.

36. Apparatus according to claim 33, further comprising means for analyzing data stored in the storage device and for producing and updating the neural network database file.