Systems and methods for dynamic detection and prevention of electronic fraud

• US 7,089,592 B2
• Filed: 03/15/2001
• Issued: 08/08/2006
• Est. Priority Date: 03/15/2001
• Status: Expired due to Term

• Current Assignee: Brighterion Inc. (MasterCard Incorporated)
• Original Assignee: Brighterion Inc. (MasterCard Incorporated)
• Inventors: Adjaoute, Akli
• Primary Examiner(s): Song, Hosuk
• Assistant Examiner(s): Son, Linh L D
• Application Number:
  US09/810,313
• Publication Number:
  US 20020133721A1
• Time in Patent Office: 1,972 Days
• Field of Search: 713/201, 706/15, 706/45, 706/20, 706/925, 706/1, 706/922, 706/2, 726/25, 726/26, 707/1, 707/3, 707/6, 717100-167, 379/145, 379/189, 455/410
• US Class Current: 726/25
• CPC Class Codes: G06Q 20/04 Payment circuits
• 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: H04L 2463/102 applying security measure for e-commerce
• CPC Class Codes: H04L 63/0407 wherein the identity of one or more communicating identities is hidden cryptographic mechanisms or cryptographic arrangements for anonymous credentials or for identity based cryptographic systems H04L9/00
• CPC Class Codes: H04L 63/1408 by monitoring network traffic monitoring network traffic per se H04L43/00

Abstract

The present invention provides systems and methods for dynamic detection and prevention of electronic fraud and network intrusion using an integrated set of intelligent technologies. The intelligent technologies include neural networks, multi-agents, data mining, case-based reasoning, rule-based reasoning, fuzzy logic, constraint programming, and genetic algorithms. The systems and methods of the present invention involve a fraud detection and prevention model that successfully detects and prevents electronic fraud and network intrusion in real-time. The model is not sensitive to known or unknown different types of fraud or network intrusion attacks, and can be used to detect and prevent fraud and network intrusion across multiple networks and industries.

Claims

1. A method for detecting and preventing electronic fraud in electronic transactions between a client and a user, the method comprising:

generating a fraud detection and prevention model software component for using a plurality of intelligent technologies to determine whether information sent by the user to the client associated with a new electronic transaction is fraudulent, wherein the model software component is trained on a database of past electronic transactions provided by the client;

querying the model software component with a current electronic transaction to determine whether information sent by the user to the client associated with the current electronic transaction is fraudulent; and

updating the model software component with the current electronic transaction, wherein the fraud detection and prevention model software component comprises a plurality of sub-models, each sub-model implementing an intelligent technology to determine whether the electronic transaction is fraudulent, wherein the plurality of sub-models respectively implement neural network technology, rule-based reasoning technology, data mining technology, and case-based reasoning technology, wherein generating the fraud detection and prevention model software component comprises using a model training interface to select which sub-models are to be included in the fraud detection and prevention model software component, wherein querying the model software component with a current electronic transaction to determine whether information sent by the user to the client associated with the current electronic transaction is fraudulent comprises providing the information as input to a binary file and running the binary file to generate a binary output decision on whether the electronic transaction is fraudulent or not, wherein running the binary file to generate the output decision on whether the electronic transaction is fraudulent comprises running the plurality of sub-models to generate a plurality of sub-model decisions and combining the plurality of sub-model decisions to generate the output decision, and wherein combining the plurality of sub-model decisions to generate the output decision comprises assigning a vote to each sub-model decision and generating the output decision based on the majority of votes determining whether the electronic transaction is fraudulent or not.

2. The method of claim 1, wherein the electronic transactions comprise web-based transactions and transactions conducted over wireless networks with the user of cellular phones.

3. The method of claim 1, wherein the plurality of intelligent technologies further comprise multi-agents, fuzzy logic, constraint programming, and genetic algorithms.

4. The method of claim 1, wherein training the model software component on a database of past electronic transactions provided by the client comprises training the plurality of sub-models and creating the binary file for implementing the plurality of sub-models.

5. The method of claim 1, wherein the database of past electronic transactions comprises a plurality of tables, wherein each table comprises a plurality of data fields and data records associated with a plurality of electronic transactions.

7. The method of claim 1, wherein combining the plurality of sub-model decisions to generate the output decision comprises providing a plurality of meta-rules to determine how the sub-model decisions are combined to generate the output decision.

8. The method of claim 1, wherein updating the model software component with the current electronic transaction comprises updating the binary file without retraining the model.

9. The method of claim 8, wherein updating the model software component with the current electronic transaction further comprises updating the database with the current electronic transaction and retraining the model to generate a new binary file.

6. A method for detecting and preventing electronic fraud in electronic transactions between a client and a user, the method comprising:

generating a fraud detection and prevention model software component for using a plurality of intelligent technologies to determine whether information sent by the user to the client associated with a new electronic transaction is fraudulent, wherein the model software component is trained on a database of past electronic transactions provided by the client;

querying the model software component with a current electronic transaction to determine whether information sent by the user to the client associated with the current electronic transaction is fraudulent; and

updating the model software component with the current electronic transaction, wherein the fraud detection and prevention model software component comprises a plurality of sub-models, each sub-model implementing an intelligent technology to determine whether the electronic transaction is fraudulent, wherein the plurality of sub-models respectively implement neural network technology, rule-based reasoning technology, data mining technology, and case-based reasoning technology, wherein generating the fraud detection and prevention model software component comprises using a model training interface to select which sub-models are to be included in the fraud detection and prevention model software component, wherein querying the model software component with a current electronic transaction to determine whether information sent by the user to the client associated with the current electronic transaction is fraudulent comprises providing the information as input to a binary file and running the binary file to generate a binary output decision on whether the electronic transaction is fraudulent or not, wherein running the binary file to generate the output decision on whether the electronic transaction is fraudulent comprises running the plurality of sub-models to generate a plurality of sub-model decisions and combining the plurality of sub-model decisions to generate the output decision, and wherein combining the plurality of sub-model decisions to generate the output decision comprises assigning a weighted vote to each one of the sub-models, wherein the weighted vote is assigned to prioritize the sub-model decisions, and generating the output decision based on the highest number of votes determining whether the electronic transaction is fraudulent or not.