Prioritizing transactions in a transaction queue

A method for prioritizing a plurality of transactions in a transaction queue, the method comprising:

• receiving, at a computer, a request for a manual review for each of the plurality of transactions;
• determining, by the computer, a probability of fraud for each transaction by comparing one or more characteristics of the transaction with corresponding characteristics of one or more previous transactions stored in a database of previous transactions;
• determining, by the computer, a first cost of fraud for each transaction for accepting the transaction, without performing the manual review, using a first function that outputs the first cost of fraud based on the probability of fraud for the transaction and a liability of a merchant conducting the transaction;
• determining, by the computer, a second cost of fraud for each transaction for performing the manual review using a second function that outputs the second cost of fraud based on the probability of fraud for the transaction and a cost of performing the manual review;
• calculating, by the computer, a difference between the first cost of fraud and the second cost of fraud for each transaction;
• determining, by the computer, a priority level for each transaction that increases as the difference increases so that transactions for which the difference is larger have a higher priority level than the transactions for which the difference is smaller;
• and
• positioning, by the computer, each transaction in the transaction queue based on the priority level of the transaction so that each transaction is scheduled to receive the manual review in order of the priority level of the transaction.

A computer program product for prioritizing a plurality of transactions in a transaction queue, the computer program product comprising:

• a non-transitory computer-readable storage medium;
• and
• instructions stored on the non-transitory computer-readable storage medium that, when executed by a processor, cause the processor to:
• receive a request for a manual review for each of the plurality of transactions;
• determine a probability of fraud for each transaction by comparing one or more characteristics of the transaction with corresponding characteristics of one or more previous transactions stored in a database of previous transactions;
• determine a first cost of fraud for each transaction for accepting the transaction, without performing the manual review, using a first function that outputs the first cost of fraud based on the probability of fraud for the transaction and a liability of a merchant conducting the transaction;
• determine a second cost of fraud for each transaction for performing the manual review using a second function that outputs the second cost of fraud based on the probability of fraud for the transaction and a cost of performing the manual review;
• calculate a difference between the first cost of fraud and the second cost of fraud for each transaction;
• determine a priority level for each transaction that increases as the difference increases so that transactions for which the difference is larger have a higher priority level than the transactions for which the difference is smaller;
• and
• position each transaction in the transaction queue based on the priority level of the transaction so that each transaction is scheduled to receive the manual review in order of the priority level of the transaction.