Detecting fraudulent mobile money transactions

1. A method for detecting a fraudulent mobile money transaction, the method comprising:

determining a reputation score for at least one customer based on historical data;

determining multiple patterns related to fraudulent mobile money transaction activities by using the reputation score and the historical data;

automatically detecting at least one of the patterns in mobile money transactions incoming from at least one other customer; and

classifying the at least one detected pattern as a fraudulent mobile money transaction or a genuine mobile money transaction;

wherein at least one of the steps is carried out by a computer device.

2. The method of claim 1 , comprising:

generating a risk score associated with the classified pattern.

3. The method of claim 1 , comprising:

incorporating domain expert knowledge to reinforce classifying the at least one detected pattern as fraudulent or genuine.

4. The method of claim 1 , wherein said classifying the at least one detected pattern comprises denoting each fraudulent mobile money transaction with a level of severity based on a determined confidence of the detected pattern classified as a fraudulent mobile money transaction and monetary impact involved with the fraudulent mobile money transaction.

5. The method of claim 1 , comprising:

storing the reputation score in a database.

6. The method of claim 1 , wherein said determining the reputation score comprises manually assigning the reputation score.

7. The method of claim 1 , wherein said determining the reputation score comprises obtaining the reputation score from an oracle.

8. The method of claim 1 , wherein said determining the reputation score comprises employing a learning scheme to learn the reputation score.

9. The method of claim 1 , wherein said determining multiple patterns comprises categorizing money laundering patterns along multiple dimensions.

10. The method of claim 9 , wherein said categorizing money laundering patterns along multiple dimensions comprises categorizing patterns relating to placement, layering, and integration activities.

11. The method of claim 1 , wherein said determining multiple patterns relating to fraudulent mobile money transaction activities comprises determining patterns based on non-network related features, network-only related features, and both non-network and network-only related features.

12. The method of claim 1 , comprising:

storing the patterns in a database.

13. An article of manufacture comprising a computer readable storage medium having computer readable instructions tangibly embodied thereon which, when implemented, cause a computer to carry out a plurality of method steps comprising:

determining a reputation score for at least one customer based on historical data;

determining multiple patterns related to fraudulent mobile money transaction activities by using the reputation score and the historical data;

automatically detecting at least one of the patterns in mobile money transactions incoming from at least one other customer; and

classifying the at least one detected pattern as a fraudulent mobile money transaction or a genuine mobile money transaction.

14. The article of manufacture of claim 13 , wherein the method steps comprise:

incorporating domain expert knowledge to reinforce classifying the at least one detected pattern as fraudulent or genuine.

15. The article of manufacture of claim 13 , wherein the method steps comprise:

denoting each fraudulent mobile money transaction with a level of severity based on a determined confidence of the detected pattern classified as a fraudulent mobile money transaction and monetary impact involved with the fraudulent mobile money transaction.

16. The article of manufacture of claim 13 , wherein determining the reputation score comprises employing a learning scheme to learn the reputation score.

17. A system for detecting a fraudulent mobile money transaction, comprising:

a memory;

at least one processor coupled to the memory; and

at least one distinct software module, each distinct software module being embodied on a tangible computer-readable medium, the at least one distinct software module comprising:

a reputation score engine module, executing on the processor, for determining a reputation score for at least one customer based on historical data;

a pattern discovery engine module, executing on the processor, for determining multiple patterns related to fraudulent mobile money transaction activities by using the reputation score and the historical data; and

a transaction workflow engine module, executing on the processor, for:

automatically detecting at least one of the patterns in mobile money transactions incoming from at least one other customer; and

classifying the at least one detected pattern as a fraudulent mobile money transaction or a genuine mobile money transaction.

18. The system of claim 17 , further comprising:

a decision support system module, executing on the processor, for facilitating a human agent to take at least one of a corrective and a precautionary action based on the at least one detected pattern.

19. The system of claim 17 , wherein the transaction workflow engine module denotes each fraudulent mobile money transaction with a level of severity based on a determined confidence of the detected pattern classified as a fraudulent mobile money transaction and monetary impact involved with the fraudulent mobile money transaction.

20. The system of claim 17 , wherein the reputation score engine module employs a learning scheme to learn the reputation score.