Automated field-mapping of account names for form population
A system for account mapping includes functionality for obtaining more than one labeled accounts labeled by more than one accountant; pre-processing more than one labeled accounts using natural language processing, using the more than one pre-processed labeled accounts to train an account mapping model that performs multinomial classification; receiving an account name from an accounting application where the account name includes a text label for an account included in a chart of accounts; generating an account mapping by applying the account mapping model to the account name, where the account mapping includes a type of the account, a sub-type of the account, a code, and a series associated with an accounting form; returning the account mapping to the accounting application through an Application Programming Interface (API); and receiving a corrected account mapping from an accountant and using the corrected account mapping as a new text label to incrementally update the account mapping model.
1. A method, comprising:
obtaining a plurality of labeled accounts labeled by a plurality of accountants;
pre-processing the plurality of labeled accounts using natural language processing;
training, using the plurality of pre-processed labeled accounts and by a computing system, an account mapping model that performs multinomial classification;
receiving, from an accounting application and by the computing system, an account name, wherein the account name is received through an application programming interface (API), and wherein the account name comprises a text label for an account included in a chart of accounts;
generating an account mapping by the computing system executing the account mapping model using the account name, wherein generating the account mapping comprises:
performing a first level classification to identify the account type of the account,
performing a second level classification to identify a sub-type of the account, wherein the sub-type is hierarchically located under the account type, and
applying a mapping to the sub-type to obtain a code and a series associated with an accounting form, wherein the code and series identifies a destination in the accounting form;
returning, by the computing system, the account mapping to the accounting application through the API;
receiving, from a client device, a correction to the account mapping; and
retraining, by the computing system, the account mapping model using upsampling based on the correction.
2. The method of claim 1 , wherein the natural language processing includes application of a bag of words model.
3. The method of claim 1 , wherein the multinomial classification is based on naïve Bayes.
4. The method of claim 1 , wherein the account name is received from the client device in a trial balance report.
5. The method of claim 4 , wherein the trial balance report includes one or more additional account names, wherein one or more additional account mappings are identified by applying the account mapping model to each of the one or more additional account names, and wherein each of the one or more additional account mappings are returned to the client device.
6. A non-transitory computer readable medium storing instructions, the instructions, when executed by a computer processor, comprising functionality for:
obtaining a plurality of labeled accounts labeled by a plurality of accountants;
pre-processing the plurality of labeled accounts using natural language processing;
training, using the plurality of pre-processed labeled accounts, an account mapping model that performs multinomial classification;
receiving, from an accounting application, an account name, wherein the account name is received through an application programming interface (API), and wherein the account name comprises a text label for an account included in a chart of accounts;
generating an account mapping by applying the account mapping model to the account name, wherein the generating the account mapping comprises:
performing a first level classification to identify the account type of the account,
performing a second level classification to identify a sub-type of the account, wherein the sub-type is hierarchically located under the account type, and
applying a mapping to the sub-type to obtain a code and a series associated with an accounting form, wherein the code and series identifies a destination in the accounting form;
returning the account mapping to the accounting application through the API;
receiving, from a client device, a correction to the account mapping; and
retraining the account mapping model using upsampling based on the correction.
7. The non-transitory computer readable medium of claim 6 , wherein the natural language processing includes application of a bag of words model.
8. The non-transitory computer readable medium of claim 6 , wherein the multinomial classification is based on naïve Bayes.
9. The non-transitory computer readable medium of claim 6 , wherein the account name is received from the client device in a trial balance report.
10. The non-transitory computer readable medium of claim 9 , wherein the trial balance report includes one or more additional account names, wherein one or more additional account mappings are identified by applying the account mapping model to each of the one or more additional account names, and wherein each of the one or more additional account mappings are returned to the client device.