Transaction counterpart identification
Provided are systems and methods for identifying a counterpart entity of a transaction from a transaction string. In one example, a method may include identifying a transaction string from an account of a target entity, the transaction string comprising a record of a payment that changed a balance of the account, executing a machine learning model based on the transaction string to determine a counterpart entity of the transaction with respect to the target entity, wherein, during execution, the machine learning model receives the transaction string as input and outputs an identifier of the counterpart entity, generating a data structure comprising a first field that stores the transaction string and a second field that stores an identifier of the counterpart entity, and storing the data structure within a file.
1. A computing system comprising:
a memory; and
a processor configured to
train a machine learning model to identify counterpart entities based on historical mappings between transaction strings and sub-strings corresponding to names of the counterpart entities,
identify a transaction string from an account of a target entity, the transaction string comprising a record of a transaction that changed a balance of the account of the target entity without necessarily identifying a counterpart entity of the target entity of the transaction, wherein the counterpart entity is an owner of an account on an opposite side of the transaction with respect to the target entity,
execute a machine learning model on the identified transaction string to determine a name of the counterpart entity of the transaction, wherein the machine learning model maps one or more words in the transaction string to the name of the counterpart entity based on learned relationships included in the historical mappings between the transaction strings and sub-strings corresponding to the names of the counterpart entities,
generate a data structure comprising a first field that stores the transaction string and a second field that stores the name of the counterpart entity, and
store the data structure within a file.
2. The computing system of claim 1 , wherein the processor is configured to map, via the machine learning model, a combination of sub-strings within the transaction string to the name of the counterpart entity during execution of the machine learning model.
3. The computing system of claim 1 , wherein the processor is configured to map, via the machine learning model, a format of the transaction string to the name of the counterpart entity during execution of the machine learning model.
4. The computing system of claim 1 , wherein the processor is configured to map, via the machine learning model, an entirety of the transaction string to the name of the counterpart entity during execution of the machine learning model.
5. The computing system of claim 1 , wherein the processor is configured to insert the transaction string into a first column of the data structure and insert the name of the counterpart entity into a second column of the data structure.
6. The computing system of claim 5 , wherein the processor is further configured to execute a second machine learning model on the first and second columns of the data structure, including the transaction string and the name of the counterpart entity, respectively, and output a predictive result based thereon.
7. The computing system of claim 1 , wherein the machine learning model comprises a neural network that performs named entity recognition which classifies string values into the name of the counterpart entity.
8. The computing system of claim 1 , wherein the processor is further configured to retrain the machine learning model based on a mapping of the transaction string to the name of the counterpart entity.
9. A method comprising:
training a machine learning model to identify counterpart entities based on historical mappings between transaction strings and sub-strings corresponding to names of the counterpart entities;
identifying a transaction string from an account of a target entity, the transaction string comprising a record of a transaction that changed a balance of the account of the target entity without necessarily identifying a counterpart entity of the transaction, wherein the counterpart entity is an owner of an account on an opposite side of the transaction with respect to the target entity;
executing a machine learning model on the identified transaction string to determine a name of the counterpart entity of the transaction, wherein the machine learning model maps one or more words in the transaction string to the name of the counterpart entity based on learned relationships included in the historical mappings between the transaction strings and sub-strings corresponding to the names of the counterpart entities;
generating a data structure comprising a first field that stores the transaction string and a second field that stores the name of the counterpart entity; and
storing the data structure within a file.
10. The method of claim 9 , wherein the executing comprises mapping, via the machine learning model, a combination of sub-strings within the transaction string to the name of the counterpart entity during execution of the machine learning model.
11. The method of claim 9 , wherein the executing comprises mapping, via the machine learning model, a format of the transaction string to the name of the counterpart entity during execution of the machine learning model.
12. The method of claim 9 , wherein the executing comprises mapping, via the machine learning model, an entirety of the transaction string to the name of the counterpart entity during execution of the machine learning model.
13. The method of claim 9 , wherein the generating comprises inserting the transaction string into a first column of the data structure and inserting the name of the counterpart entity into a second column of the data structure.
14. The method of claim 13 , wherein the method further comprises executing a second machine learning model on the first and second columns of the data structure including the transaction string and the name of the counterpart entity, respectively, and outputting a predictive result based thereon.
15. The method of claim 9 , wherein the machine learning model comprises a neural network that performs named entity recognition which classifies string values into the name of the counterpart entity.
16. The method of claim 9 , wherein the method further comprises retraining the machine learning model based on a mapping of the transaction string to the name of the counterpart entity.
17. A non-transitory computer-readable medium comprising instructions which when executed by a processor cause a computer to perform a method comprising:
training a machine learning model to identify counterpart entities based on historical mappings between transaction strings and sub-strings corresponding to the names of the counterpart entities;
identifying a transaction string from an account of a target entity, the transaction string comprising a record of a transaction that changed a balance of the account of the target entity without necessarily identifying a counterpart entity of the transaction, wherein the counterpart entity is an owner of an account on an opposite side of the transaction with respect to the target entity;
executing a machine learning model on the identified transaction string to determine a name of the counterpart entity of the transaction, wherein the machine learning model maps one or more words in the transaction string to the name of the counterpart entity based on learned relationships included in the historical mappings between the transaction strings and sub-strings corresponding to the names of the counterpart entities;
generating a data structure comprising a first field that stores the transaction string and a second field that stores the name of the counterpart entity; and
storing the data structure within a file.
18. The non-transitory computer-readable medium of claim 17 , wherein the executing comprises mapping, via the machine learning model, a combination of sub-strings within the transaction string to the name of the counterpart entity during execution of the machine learning model.
19. The non-transitory computer-readable medium of claim 17 , wherein the executing comprises mapping, via the machine learning model, a format of the transaction string to the name of the counterpart entity during execution of the machine learning model.
20. The non-transitory computer-readable medium of claim 17 , wherein the executing comprises mapping, via the machine learning model, an entirety of the transaction string to the name of the counterpart entity during execution of the machine learning model.