Determining a temporary transaction limit

1. A system, comprising:

a processor configured to:

receive a transaction request message, wherein the transaction request message includes a total transaction cost associated with a requested transaction and identifying information associated with a user, wherein the user is associated with a plurality of historical transaction costs;

dynamically determine a predetermined limit associated with the user, wherein to dynamically determine the predetermined limit comprises to:

rank, by size, the plurality of historical transaction costs associated with the user;

determine an outlier condition;

determine one or more historical transaction costs associated with the user that meet the outlier condition from the plurality of historical transaction costs associated with the user; and

determine the predetermined limit associated with the user based at least in part on a maximum value from the plurality of historical transaction costs associated with the user excluding the one or more historical transaction costs associated with the user that meet the outlier condition;

determine the total transaction cost is greater than the predetermined limit;

obtain a plurality of clustered classifications that is generated from historical data of a plurality of users based at least in part on clustering at least a portion of the historical data of the plurality of users;

select a clustered classification corresponding to the user from the plurality of clustered classifications based at least in part on the plurality of historical transaction costs associated with the user, including to:

apply a user score derived based on the plurality of historical transaction costs associated with the user to determine the selected clustered classification;

determine a dynamic quota corresponding to the clustered classification for the user using a predetermined mapping rule, wherein a first clustered classification is related by the predetermined mapping rule to a first dynamic quota and a second clustered classification is related by the predetermined mapping rule to a second dynamic quota, wherein the first dynamic quota is different from the second dynamic quota;

determine to approve the transaction request message based at least in part on determining that the total transaction cost is less than a temporary transaction limit, wherein the temporary transaction limit comprises a combination of the predetermined limit and the dynamic quota; and

in response to the determination to approve the transaction request message, complete the requested transaction; and

a memory coupled to the processor and configured to provide the processor with instructions.

2. The system of claim 1 , wherein the processor is further configured to generate the plurality of clustered classifications from the historical data of the plurality of users by clustering on at least a portion of the historical data of the plurality of users along one or more measures.

3. The system of claim 2 , wherein the one or more measures comprise one or more of the following: most recent spending, spending frequency, and amount of spending.

4. The system of claim 1 , wherein to select the clustered classification from the plurality of clustered classifications based at least in part on the plurality of historical transaction costs associated with the user comprises to:

convert the plurality of historical transaction costs associated with the user into values corresponding to one or more measures; and

compare the values corresponding to the one or more measures to ranges of values corresponding to the one or more measures associated with the clustered classification.

5. The system of claim 1 , wherein the temporary transaction limit comprises a sum of the predetermined limit and the dynamic quota.

6. A method, comprising:

receiving a transaction request message, wherein the transaction request message includes a total transaction cost associated with a requested transaction and identifying information associated with a user, wherein the user is associated with a plurality of historical transaction costs;

dynamically determining a predetermined limit associated with the user, wherein dynamically determining the predetermined limit comprises:

ranking, by size, the plurality of historical transaction costs associated with the user;

determining an outlier condition;

determining one or more historical transaction costs associated with the user that meet the outlier condition from the plurality of historical transaction costs associated with the user; and

determining the predetermined limit associated with the user based at least in part on a maximum value from the plurality of historical transaction costs associated with the user excluding the one or more historical transaction costs associated with the user that meet the outlier condition;

determining the total transaction cost is greater than the predetermined limit;

obtaining a plurality of clustered classifications that is generated from historical data of a plurality of users based at least in part on clustering at least a portion of the historical data of the plurality of users;

selecting, using one or more processors, a clustered classification corresponding to the user from the plurality of clustered classifications based at least in part on the plurality of historical transaction costs associated with the user, including:

applying a user score derived based on the plurality of historical transaction costs associated with the user to determine the selected clustered classification;

determining a dynamic quota corresponding to the clustered classification for the user using a predetermined mapping rule, wherein a first clustered classification is related by the predetermined mapping rule to a first dynamic quota and a second clustered classification is related by the predetermined mapping rule to a second dynamic quota, wherein the first dynamic quota is different from the second dynamic quota;

determining to approve the transaction request message based at least in part on determining that the total transaction cost is less than a temporary transaction limit, wherein the temporary transaction limit comprises a combination of the predetermined limit and the dynamic quota; and

in response to the determination to approve the transaction request message, completing the requested transaction.

7. The method of claim 6 , further comprising generating the plurality of clustered classifications from the historical data of the plurality of users by clustering on at least a portion of the historical data of the plurality of users along one or more measures.

8. The method of claim 7 , wherein the one or more measures comprise one or more of the following: most recent spending, spending frequency, and amount of spending.

9. The method of claim 6 , wherein selecting the clustered classification from the plurality of clustered classifications based at least in part on the plurality of historical transaction costs associated with the user comprises:

converting the plurality of historical transaction costs associated with the user into values corresponding to one or more measures; and

comparing the values corresponding to the one or more measures to ranges of values corresponding to the one or more measures associated with the clustered classification.

10. The method of claim 6 , wherein the temporary transaction limit comprises a sum of the predetermined limit and the dynamic quota.

11. A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

receiving a transaction request message, wherein the transaction request message includes a total transaction cost associated with a requested transaction and identifying information associated with a user, wherein the user is associated with a plurality of historical transaction costs;

dynamically determining a predetermined limit associated with the user, wherein dynamically determining the predetermined limit comprises:

ranking, by size, the plurality of historical transaction costs associated with the user;

determining an outlier condition;

determining one or more historical transaction costs associated with the user that meet the outlier condition from the plurality of historical transaction costs associated with the user; and

determining the predetermined limit associated with the user based at least in part on a maximum value from the plurality of historical transaction costs associated with the user excluding the one or more historical transaction costs associated with the user that meet the outlier condition;

determining the total transaction cost is greater than the predetermined limit;

obtaining a plurality of clustered classifications that is generated from historical data of a plurality of users based at least in part on clustering at least a portion of the historical data of the plurality of users;

selecting a clustered classification corresponding to the user from the plurality of clustered classifications based at least in part on the plurality of historical transaction costs associated with the user, including:

applying a user score derived based on the plurality of historical transaction costs associated with the user to determine the selected clustered classification;

determining a dynamic quota corresponding to the clustered classification for the user using a predetermined mapping rule, wherein a first clustered classification is related by the predetermined mapping rule to a first dynamic quota and a second clustered classification is related by the predetermined mapping rule to a second dynamic quota, wherein the first dynamic quota is different from the second dynamic quota;

determining to approve the transaction request message based at least in part on determining that the total transaction cost is less than a temporary transaction limit, wherein the temporary transaction limit comprises a combination of the predetermined limit and the dynamic quota; and

in response to the determination to approve the transaction request message, completing the requested transaction.

12. The system of claim 1 , wherein the outlier condition is determined based at least in part on |d i −d i-k |>C,i=k+1, . . . , n,

where i indicates an ith historical transaction cost associated with the user, d 1 , d 2 , . . . , d n comprises n number of historical transaction costs associated with the user ranked by size, C is a threshold value, and k is a preset distance.

13. The system of claim 1 , wherein the historical data of the plurality of users comprises at least partially structured historical data of the plurality of users and wherein the processor is further configured to:

integrate raw historical data of users from a plurality of systems;

generate the at least partially structured historical data of the plurality of users by structuring the raw historical data of users; and

store the at least partially structured historical data of the plurality of users in one or more databases.