Techniques for utilizing a predictive model to cache processing data
Embodiments of the invention are directed to systems and methods for utilizing a cache to store historical transaction data. A predictive model may be trained to identify particular identifiers associated with historical data that is likely to be utilized on a particular date and/or within a particular time period. The historical data corresponding to these identifiers may be stored in a cache of the processing computer. Subsequently, an authorization request message may be received that includes an identifier. The processing computer may utilize the identifier to retrieve historical transaction data from the cache. The retrieved data may be utilized to perform any suitable operation. By predicting the data that will be needed to perform these operations, and preemptively store such data in a cache, the latency associated with subsequent processing may be reduced and the performance of the system as a whole improved.
1. A computer-implemented method, comprising:
periodically updating, by a computing device, a cache of the computing device by:
identifying, by the computing device utilizing a predictive model, one or more accounts from a plurality of accounts, the one or more accounts being identified by the predictive model as being likely to be utilized to conduct one or more transactions within a future time period, the predictive model being trained based at least in part on historical transaction data associated with the plurality of accounts, the historical transaction data identifying a number of the plurality of accounts that have historically been utilized within a particular historical time period, wherein the number of the plurality of accounts are determined to have conducted over a threshold percentage of transactions occurring within the particular historical time period, wherein the predictive model is trained utilizing unsupervised machine learning techniques;
obtaining, by the computing device from a data store, a portion of the historical transaction data, the portion of the historical transaction data corresponding to transactions conducted utilizing the one or more accounts;
based on identifying the one or more accounts as being likely to be utilized within the future time period, storing, at the cache of the computing device, the portion of the historical transaction data; and
deleting, from the cache of the computing device, another portion of the historical transaction data;
receiving, by the computing device over a communications network, an authorization request message for a transaction associated with an account;
determining whether historical transaction data associated with the account is stored in the cache;
when the historical transaction data is stored in the cache:
retrieving, by the computing device, the historical transaction data associated with the account from the cache;
in response to receiving the authorization request message, utilizing, by the computing device, the historical transaction data retrieved from the cache to calculate a risk score for the authorization request message and transmit an authorization response comprising the risk score; and
the transaction is completed and processed based on the authorization response; and
when the historical transaction data is not stored in the cache:
retrieving, by the computing device, the historical transaction data associated with the account from a remote historical transaction database;
in response to receiving the authorization request message, utilizing, by the computing device, the historical transaction data retrieved from the historical transaction database to calculate a risk score for the authorization request message and transmit an authorization response comprising the risk score over the communications network; and
the transaction is completed and processed based on the authorization response.
2. The computer-implemented method of claim 1 , further comprising updating the historical transaction data with transaction data associated with authorization request message, wherein the predictive model is updated based at least in part on the historical transaction data as updated.
3. The computer-implemented method of claim 1 , wherein identifying the one or more accounts, obtaining the portion of the historical transaction data, and storing the portion of the historical transaction data, are performed in response to receiving the authorization request message.
4. The computer-implemented method of claim 1 , further comprising updating the historical transaction data with transaction data of the authorization request message, wherein the one or more accounts are identified from the plurality of accounts subsequent to the historical transaction data being updated with the transaction data.
5. The computer-implemented method of claim 1 , wherein the computing device is a processing network computer.
6. The method of claim 1 , further comprising training the predictive model on the historical transaction data to utilize a season, holiday information corresponding to transaction dates, and demographic information associated with an account holder to identify the accounts that are likely to be utilized within the future time period.
7. A processing network computer, comprising:
a processor; and
a computer readable medium, the computer readable medium comprising code, executable by the processor, for implementing a method comprising:
periodically updating a cache of the processing network computer by:
identifying, utilizing a predictive model, one or more accounts from a plurality of accounts, the one or more accounts being identified by the predictive model as being likely to be utilized to conduct one or more transactions within a future time period, the predictive model being trained based at least in part on historical transaction data associated with the plurality of accounts, the historical transaction data identifying a number of the plurality of accounts that have historically been utilized within a particular historical time period, wherein the number of the plurality of accounts are determined to have conducted over a threshold percentage of transactions occurring within the particular historical time period, wherein the predictive model is trained utilizing unsupervised machine learning techniques;
obtaining, from a data store, a portion of the historical transaction data, the portion of the historical transaction data corresponding to transactions conducted utilizing the one or more accounts;
based on identifying the one or more accounts as being likely to be utilized within the future time period, storing the portion of the historical transaction data within the cache of the processing network computer; and
deleting, from the cache of the processing network computer, another portion of the historical transaction data;
receiving, over a communications network, an authorization request message for a transaction associated with an account;
determining whether historical transaction data associated with the account is stored in the cache;
when the historical transaction data is stored in the cache:
retrieving the historical transaction data associated with the account from the cache;
in response to receiving the authorization request message, utilizing the historical transaction data retrieved from the cache to calculate a risk score for the authorization request message and transmit an authorization response comprising the risk score; and
the transaction is completed and processed based on the authorization response; and
when the historical transaction data is not stored in the cache:
retrieving the historical transaction data associated with the account from a remote historical transaction database;
in response to receiving the authorization request message, utilizing the historical transaction data retrieved from the historical transaction database to calculate a risk score for the authorization request message and transmit an authorization response comprising the risk score over the communications network; and
the transaction is completed and processed based on the authorization response.
8. The processing network computer of claim 7 , the method further comprising updating the historical transaction data with transaction data associated with the authorization request message, wherein the predictive model is updated based at least in part on the historical transaction data as updated.
9. The processing network computer of claim 7 , wherein identifying the one or more accounts, obtaining the portion of the historical transaction data, and storing the portion of the historical transaction data, are performed in response to receiving the authorization request message.
10. The processing network computer of claim 7 , the method further comprising updating the historical transaction data with transaction data of the authorization request message, wherein the one or more accounts are identified from the plurality of accounts subsequent to the historical transaction data being updated with the transaction data.
11. The processing network computer of claim 7 , the method further comprising training the predictive model on the historical transaction data to utilize a season, holiday information corresponding to transaction dates, and demographic information associated with an account holder to identify the accounts that are likely to be utilized within the future time period.