Identifying potentially risky transactions

1. A computer-implemented method, comprising:

determining, by a processor, a plurality of characteristics of a new transaction;

processing, by the processor, the plurality of characteristics of the new transaction using a risk assessment model, the risk assessment model generated using an iterative training process that processes, over time, historical transactions to train the risk assessment model based on characteristics of historical transactions and associated risk assessment analysis of the historical transactions;

transforming, by the processor, each of the plurality of characteristics of the new transaction into a corresponding numerical value; and

generating, by the processor, a risk assessment indicator for the new transaction based on the corresponding numerical value for each of the plurality of characteristics of the new transaction, the risk assessment indicator indicating a degree of risk of the new transaction,

wherein processing, by the processor, the plurality of characteristics of the new transaction using the risk assessment model comprises:

identifying, based on the plurality of characteristics of the new transaction, a set of classifiers, wherein each classifier of the set of classifiers corresponds to at least one of the characteristics of the new transaction and represents an aspect of risk associated with the new transaction, and wherein, using each classifier, the corresponding numerical value for each of the plurality of characteristics of the new transaction is generated; and

determining, based on the set of classifiers, a plurality of ensemble values for the new transaction,

and wherein generating the risk assessment indicator comprises:

using each classifier from among the set of classifiers for detecting a risk probability of one of the characteristics of the new transaction, wherein the risk probability is represented by one of the ensemble values;

determining, from the plurality of ensemble values, an ensemble maximum value and an ensemble minimum value, wherein such values comprise, respectively, a maximum risk probability and a minimum risk probability with respect to the set of classifiers; and

averaging the maximum risk probability and the minimum risk probability.

2. The computer-implemented method of claim 1 , further comprising generating the risk assessment model using the iterative training process, the iterative training process comprising applying machine learning techniques that train the risk assessment model over time and based on the historical transactions.

3. The computer-implemented method of claim 1 , wherein processing, by the processor, the plurality of characteristics of the new transaction using the risk assessment model comprises applying one or more of the following:

fuzzy matching techniques; and

entity disambiguation techniques.

4. The computer-implemented method of claim 1 , wherein generating, by the processor, the risk assessment indicator further comprises computing an average of the plurality of ensemble values.

5. The computer-implemented method of claim 1 , wherein the risk assessment indicator comprises a numeric value.

6. The computer-implemented method of claim 1 , further comprising updating the risk assessment model based on the new transaction.

7. The computer-implemented method of claim 1 , further comprising:

comparing a value of the risk assessment indicator for the new transaction to a first risk assessment threshold value;

and automatically escalating, in response to determining that the value of the risk assessment indicator is greater than or equal to the first risk assessment threshold value, the new transaction to a human review process in which the new transaction is assigned to a human reviewer for review of the new transaction.

8. The computer-implemented method of claim 7 , further comprising:

comparing the value of the risk assessment indicator for the new transaction to a second risk assessment threshold value; and

automatically blocking, in response to determining that the value of the risk assessment indicator is greater than or equal to the second risk assessment threshold value, the transaction.

9. The computer-implemented method of claim 1 , further comprising:

comparing a value of the risk assessment indicator for the new transaction to a risk assessment threshold value; and

automatically approving, in response to determining that the value of the risk assessment indicator is less than the risk assessment threshold value, the transaction.

10. A system comprising:

a memory storing a risk assessment model, the risk assessment model generated using a machine learning process that processes, over time, historical transactions; and

a processor, communicatively coupled to the memory, operable to perform operations comprising:

determining a plurality of characteristics of a new transaction;

processing the plurality of characteristics of the new transaction using a risk assessment model, the risk assessment model generated using an iterative training process that processes, over time, historical transactions to train the risk assessment model based on characteristics of historical transactions and associated risk assessment analysis of the historical transactions;

transforming each of the plurality of characteristics of the new transaction into a corresponding numerical value; and

generating a risk assessment indicator for the new transaction based on the corresponding numerical value for each of the plurality of characteristics of the new transaction, the risk assessment indicator indicating a degree of risk of the new transaction,

wherein the processor is further operable to process the plurality of characteristics of the new transaction using the risk assessment model by:

identifying, based on the plurality of characteristics of the new transaction, a set of classifiers, wherein each classifier of the set of classifiers corresponds to at least one of the characteristics of the new transaction and represents an aspect of risk associated with the new transaction, and wherein, using each classifier, the corresponding numerical value for each of the plurality of characteristics of the new transaction is generated;

determining, based on the set of classifiers, a plurality of ensemble values for the new transaction;

using each classifier from among the set of classifiers for detecting a risk probability of one of the characteristics of the new transaction, wherein the risk probability is represented by one of the ensemble values;

determining, from the plurality of ensemble values, an ensemble maximum value and an ensemble minimum value, wherein such values comprise, respectively, a maximum risk probability and a minimum risk probability with respect to the set of classifiers; and

averaging the maximum risk probability and the minimum risk probability.

11. The system of claim 10 , wherein the processor is further operable to generate the risk assessment model using the iterative training process, the iterative training process comprising applying machine learning techniques that train the risk assessment model over time and based on the historical transactions.

12. The system of claim 10 , wherein the processor is further operable to process the plurality of characteristics of the new transaction using the risk assessment model by applying one or more of the following:

fuzzy matching techniques; and

entity disambiguation techniques.

13. The computer-implemented method of claim 10 , wherein generating the risk assessment indicator comprises computing an average of the plurality of ensemble values.

14. The computer-implemented method of claim 10 , wherein the risk assessment indicator comprises a numeric value.

15. The system of claim 10 , wherein the processor is further operable to perform operations comprising updating the risk assessment model based on the new transaction.

16. The system of claim 10 , wherein the processor is further operable to perform operations comprising:

comparing a value of the risk assessment indicator for the new transaction to a first risk assessment threshold value; and

automatically escalating, in response to determining that the value of the risk assessment indicator is greater than or equal to the first risk assessment threshold value, the new transaction to a human review process in which the new transaction is assigned to a human reviewer for review of the new transaction.

17. The system of claim 16 , wherein the processor is further operable to perform operations comprising:

comparing the value of the risk assessment indicator for the new transaction to a second risk assessment threshold value; and

automatically blocking, in response to determining that the value of the risk assessment indicator is greater than or equal to the second risk assessment threshold value, the transaction.

18. The system of claim 10 , wherein the processor is further operable to perform operations comprising:

comparing a value of the risk assessment indicator for the new transaction to a risk assessment threshold value; and

automatically approving, in response to determining that the value of the risk assessment indicator is less than the risk assessment threshold value, the transaction.

19. A non-transitory computer readable medium storing instructions that, when executed by a processor, cause the processor to perform operations comprising:

determining a plurality of characteristics of a new transaction;

processing the plurality of characteristics of the new transaction using a risk assessment model, the risk assessment model generated using an iterative training process that processes, over time, historical transactions to train the risk assessment model based on characteristics of historical transactions and associated risk assessment analysis of the historical transactions;

transforming each of the plurality of characteristics of the new transaction into a corresponding numerical value; and

generating a risk assessment indicator for the new transaction based on the corresponding numerical value for each of the plurality of characteristics of the new transaction, the risk assessment indicator indicating a degree of risk of the new transaction,

wherein the processing the plurality of characteristics of the new transaction using the risk assessment model comprises:

identifying, based on the plurality of characteristics of the new transaction, a set of classifiers, wherein each classifier of the set of classifiers corresponds to at least one of the characteristics of the new transaction and represents an aspect of risk associated with the new transaction, and wherein, using each classifier, the corresponding numerical value for each of the plurality of characteristics of the new transaction is generated; and

determining, based on the set of classifiers, a plurality of ensemble values for the new transaction,

and wherein generating the risk assessment indicator comprises:

using each classifier from among the set of classifiers for detecting a risk probability of one of the characteristics of the new transaction, wherein the risk probability is represented by one of the ensemble values;

determining, from the plurality of ensemble values, an ensemble maximum value and an ensemble minimum value, wherein such values comprise, respectively, a maximum risk probability and a minimum risk probability with respect to the set of classifiers; and

averaging the maximum risk probability and the minimum risk probability.