Machine learning based automated electronic document processing system
Aspects of the present disclosure relate to automated document processing systems. In particular, embodiments disclosed herein provide an architecture for seamlessly integrating code representations of electronic documents generated by various different extraction systems into a single document processing system. Certain embodiments provide for generating a code representation of an electronic document based on an entity extracted from the electronic document. The code representation includes an entity representation that comprises an index indicator appended to an identifier for the entity, which may comprise a column of items. Some embodiments provide for applying a transformation to the entity representation to produce an intermediate code representation of the electronic document, wherein the transformation comprises a format for entities or a format for entity representations. Certain embodiments provide for generating an updated code representation of the electronic document based on matching indexed item values from the electronic document to the transformed entity representation.
1 . A method of automatically processing electronic documents, comprising:
generating, using a generative machine learning model, a code representation of an electronic document based on an entity comprising a row of items or a column of items of the electronic document;
generating an entity representation for the entity based on determining that the code representation corresponds to an incompatible format for an electronic document processing application, wherein the entity representation comprises an index indicator appended to an identifier for the entity;
applying a transformation to the entity representation to produce an intermediate code representation of the electronic document, wherein the transformation comprises a format for entities or a format for code representations of entities;
generating an updated code representation of the electronic document based on matching indexed item values from the electronic document to the transformed entity representation, wherein the undated code representation corresponds to a compatible format for the electronic document processing application; and
uploading the updated code representation to the electronic document processing application.
2 . The method of claim 1 , wherein the entity is extracted using an extraction machine learning model that is configured to extract entities from documents.
3 . The method of claim 1 , wherein the entity comprises a column and each respective item within the entity comprises a value for the column at a respective row.
4 . The method of claim 3 , wherein generating the updated code representation of the electronic document comprises generating an item representation comprising a given value and a given index position indicator appended to the identifier for the entity, wherein the given value is a value for the column at a row that is indicated by the given index position indicator.
5 . The method of claim 4 , wherein item values are extracted and stored in an index, wherein the given value is retrieved from the index based on matching the given index position indicator to a position in the index that contains the given value.
6 . The method of claim 1 , wherein the transformation is applied automatically based on a target format for the transformed entity representation, wherein the target format is a format that is compatible with the electronic document processing application.
7 . The method of claim 6 , further comprising automatically processing the electronic document by the electronic document processing application based on the updated code representation.
8 . The method of claim 1 , wherein the transformation is applied based on user input specifying a format for entities or a format for entity representations.
9 . The method of claim 1 , wherein the format for entity representations comprises a different identifier for the entity.
10 . A system for automatically processing electronic documents, comprising:
one or more processors; and
a memory comprising instructions that, when executed by the one or more processors, cause the system to:
generate, using a generative machine learning model, a code representation of an electronic document based on an entity comprising a row of items or a column of items of the electronic document;
generate an entity representation for the entity based on determining that the code representation corresponds to an incompatible format for an electronic document processing application, wherein the entity representation comprises an index indicator appended to an identifier for the entity;
apply a transformation to the entity representation to produce an intermediate code representation of the electronic document, wherein the transformation comprises a format for entities or a format for code representations of entities;
generate an updated code representation of the electronic document based on matching indexed item values from the electronic document to the transformed entity representation, wherein the updated code representation corresponds to a compatible format for the electronic document processing application; and
upload the updated code representation to the electronic document processing application.
11 . The system of claim 10 , wherein the entity is extracted using an extraction machine learning model that is configured to extract entities from documents.
12 . The system of claim 10 , wherein the entity comprises a column and each respective item within the entity comprises a value for the column at a respective row.
13 . The system of claim 12 , wherein generating the updated code representation of the electronic document comprises generating an item representation comprising a given value and a given index position indicator appended to the identifier for the entity, wherein the given value is a value for the column at a row that is indicated by the given index position indicator.
14 . The system of claim 13 , wherein item values are extracted and stored in an index, wherein the given value is retrieved from the index based on matching the given index position indicator to a position in the index that contains the given value.
15 . The system of claim 10 , wherein the transformation is applied automatically based on a target format for the transformed entity representation, wherein the target format is a format that is compatible with an electronic document processing system.
16 . The system of claim 10 , wherein the transformation is applied based on user input specifying a format for entities or a format for entity representations.
17 . The system of claim 10 , wherein the format for entity representations comprises a different identifier for the entity.
18 . A non-transitory computer readable storage medium comprising instructions, that when executed by one or more processors of a computing system, cause the computing system to:
generate, using a generative machine learning model, a code representation of an electronic document based on an entity comprising a row of items or a column of items of the electronic document;
generate an entity representation for the entity based on determining that the code representation corresponds to an incompatible format for an electronic document processing application, wherein the entity representation comprises an index indicator appended to an identifier for the entity;
apply a transformation to the entity representation to produce an intermediate code representation of the electronic document, wherein the transformation comprises a format for entities or a format for code representations of entities;
generate an updated code representation of the electronic document based on matching indexed item values from the electronic document to the transformed entity representation, wherein the updated code representation corresponds to a compatible format for the electronic document processing application; and
upload the updated code representation to the electronic document processing application.