System and method for detecting fraudulent transactions
View Patent ↗Some demonstrative embodiments of the invention include a system and method for detecting related transactions, e.g., fraudulent transactions by for example associating transactions and creating lists of related transactions. Other embodiments are described and claimed.
1. A method for identifying fraudulent transactions comprising:
receiving, by a processor, a set of at least one marked transaction record each transaction record in said set of marked transaction records respectively associated with a fraud score indicating a fraudulent transaction and a set of unmarked transaction records;
for each unmarked transaction record in the set of unmarked transaction records, generating, by the processor, a relation value for the unmarked transaction record with respect to at least one marked transaction record, said relation value based on an outcome of at least one equivalence function therebetween;
associating, by the processor, unmarked transaction records satisfying a marking condition with a fraud score, said fraud score based on said relation value of the unmarked transaction record with respect to the marked transaction record and its respective fraud score, thereby rendering unmarked transaction records having a fraud score as newly marked transaction records; and
repeating said steps of generating a relation value and associating unmarked transaction records with a fraud score using the marked and newly marked transaction records as the set of marked transaction records, until a termination condition is reached.
2. The method of claim 1 , further comprising after said termination condition is reached, reporting said marked transaction records as fraudulent.
3. The method of claim 2 , wherein said relation value is based on outcomes of a plurality of equivalence functions applied to unmarked transaction records with respect to said marked transaction record.
4. The method of claim 3 , wherein said transaction records each include a plurality of data fields and said plurality of equivalence functions compare at least one data field in said unmarked transaction records to corresponding at least one data field in said marked transaction record.
5. The method of claim 4 , wherein said transaction records include information selected from the set consisting of: date of transaction, time of transaction, merchant, transaction amount, IP address, account number, and cookie.
6. The method of claim 4 , wherein said data fields include an IP address, and said plurality of equivalence functions includes an equivalence function comparing an IP address associated with said unmarked transaction records with an IP address of said marked record.
7. The method of claim 4 , wherein said data fields include a transaction amount, and said plurality of equivalence functions includes an equivalence function comparing a transaction amount associated with said unmarked transaction records with a transaction amount of said marked record.
8. The method of claim 4 , wherein said data fields include a credit card number, and said plurality of equivalence functions includes an equivalence function comparing a credit card number associated with said unmarked transaction records with a credit card number of said marked record.
9. The method of claim 4 , wherein said data fields include a merchant, and said plurality of equivalence functions includes an equivalence function comparing a merchant associated with said unmarked transaction records with a merchant of said marked record.
10. The method of claim 4 , wherein said data fields include a date and a time, and said plurality of equivalence functions includes an equivalence function comparing a date and a time associated with said unmarked transaction records with a date and a time of said marked record.
11. The method of claim 4 , wherein said data fields include a cookie, and said plurality of equivalence functions includes an equivalence function comparing a cookie associated with said unmarked transaction records with a cookie of said marked record.
12. The method of claim 4 , wherein said relation value for each unmarked record comprises a sum of the outcomes of the plurality of equivalence functions applied to said unmarked transaction record with respect to said marked transaction record.
13. The method of claim 12 , wherein said marking condition is a maximum number of repetitions of the steps of providing a relation value for unmarked transaction records and marking transaction records satisfying a marking condition based on said relation value.
14. A system for identifying fraudulent transactions comprising:
a processor to receive a set of at least one marked transaction record each transaction record in said set of marked transaction records respectively associated with a fraud score indicating a fraudulent transaction and a set of unmarked transaction records, for each unmarked transaction record in the set of unmarked transaction records generate a relation value for the unmarked transaction record with respect to at least one marked transaction record, said relation value based on an outcome of at least one equivalence function therebetween, associate unmarked transaction records satisfying a marking condition with a fraud score, said fraud score based on said relation value of the unmarked transaction record with respect to the marked transaction record and its respective fraud score, thereby rendering unmarked transaction records having a fraud score as newly marked transaction records, and repeat said steps of generating a relation value and associating transaction records with a fraud score using the marked and newly marked transaction records as the set of marked transaction records, until a termination condition is reached.
15. The system of claim 14 , wherein said processor is further to report said marked transaction records as fraudulent.
16. The system of claim 15 , wherein said relation value is based on outcomes of a plurality of equivalence functions applied to unmarked transaction records with respect to said marked transaction record.
17. The system of claim 16 , wherein said transaction records each include a plurality of data fields and said plurality of equivalence functions compare at least one data field in said unmarked transaction records to corresponding at least one data field in said marked transaction record.
18. The system of claim 17 , wherein said transaction records include information selected from the set consisting of: date of transaction, time of transaction, merchant, transaction amount, IP address, account number, and cookie.
19. The system of claim 17 , wherein said data fields include an IP address, and said plurality of equivalence functions includes an equivalence function comparing an IP address associated with said unmarked transaction records with an IP address of said marked record.
20. The system of claim 17 , wherein said data fields include a transaction amount, and said plurality of equivalence functions includes an equivalence function comparing a transaction amount associated with said unmarked transaction records with a transaction amount of said marked record.
21. The system of claim 17 , wherein said data fields include a credit card number, and said plurality of equivalence functions includes an equivalence function comparing a credit card number associated with said unmarked transaction records with a credit card number of said marked record.
22. The system of claim 17 , wherein said data fields include a merchant, and said plurality of equivalence functions includes an equivalence function comparing a merchant associated with said unmarked transaction records with a merchant of said marked record.
23. The system of claim 17 , wherein said data fields include a date and a time, and said plurality of equivalence functions includes an equivalence function comparing a date and a time associated with said unmarked transaction records with a date and a time of said marked record.
24. The system of claim 17 , wherein said data fields include a cookie, and said plurality of equivalence functions includes an equivalence function comparing a cookie associated with said unmarked transaction records with a cookie of said marked record.
25. The system of claim 17 , wherein said relation value for each unmarked record comprises a sum of the outcomes of the plurality of equivalence functions applied to said unmarked transaction record with respect to said marked transaction record.
26. The system of claim 25 , wherein said marking condition is a maximum number of repetitions of the steps of providing a relation value for unmarked transaction records and said processor is to mark transaction records satisfying a marking condition based on said relation value.
27. The method of claim 1 , wherein the fraud score of said newly marked transaction record is diluted in value compared to the fraud scores of the received marked transaction records.
28. The system of claim 14 , wherein the fraud score of said newly marked transaction records is diluted in value compared to the fraud score of the received marked transaction records.