Document processing

1. A document processing system that extracts editable data from electronic documents, the system comprising:

one or more processors; and

a non-transitory data storage comprising machine-executable instructions that cause the one or more processors to:

convert a non-editable image file into a markup file,

the non-editable image file pertaining to an electronic document, and

the electronic document including at least one component document, and

the markup file preserving a format and a structure of the component document from the image file;

parse the markup file to extract plain text data of the non-editable image file;

determine styling information of the non-editable image file from the markup file;

automatically segment into sections, the plain text data by analyzing contents of the markup file according to boundary rules;

identify and extract entities automatically from the segmented sections of the plain text data,

the identifying performed using natural language processing (NLP);

link the entities using at least one of:

domain-specific ontologies,

knowledge bases, and

graphical inferences;

generate an interactive graphical user interface (GUI),

the GUI displaying content,

the content including the plain text data formatted according to the styling information,

the GUI including annotated entity relationships determined from the linked entities, and

the GUI enabling user interactions for editing one or more of the content and the annotated entity relationships; and

transmit user edits of one or more of the entity relationships, the content, the structure and the format to the knowledge bases.

2. The system of claim 1 , wherein the instructions to convert the non-editable image file into the markup file further comprise instructions that cause the one or more processors to:

identify features of textual data within the non-editable image file using optical character recognition;

determine non-textual details including document structure of the component document;

obtain markup tags associated with features of the textual data and the non-textual details; and

generate the markup file by applying the markup tags to the textual data per the document structure.

3. The system of claim 1 , wherein the instructions to automatically segment the plain text data further cause the one or more processors to:

predict whether each section of the plain text data corresponding to a page of the markup document represents a document boundary,

the predicting performed using a trained machine learning (ML) boundary model and the boundary rules.

4. The system of claim 3 , wherein the instructions to automatically segment the plain text data further cause the one or more processors to:

create a page break at a predicted boundary between the sections.

5. The system of claim 3 , wherein the instructions to automatically segment the plain text data further cause the one or more processors to:

predict whether a section of the plain text data is relevant or irrelevant to a current task using relevance rules; and

remove non-relevant sections from the plain text data.

6. The system of claim 1 , wherein the instructions to automatically identify and extract entities further cause the one or more processors to:

identify the entities in the plain text data using the domain-specific ontologies and the knowledge bases; and

categorize the entities into condition entities and evidence entities based on respective semantic type of the entities.

7. The system of claim 1 , wherein the instructions to automatically identify and extract entities further cause the one or more processors to:

assess relevance of the entities based on relevance rules that include analyzing at least a context surrounding the entities in the plain text data; and

remove from further processing, a subset of the entities based on relevance rules.

8. The system of claim 1 , wherein the instructions to link the entities further cause the one or more processors to:

identify the links from evidence entities to condition entities using domain-specific ontologies; and

discover new candidate links between the condition entities and the evidence entities by employing knowledge graphs stored in the knowledge bases.

9. The system of claim 8 , wherein the non-transitory data storage comprising instructions that further cause the one or more processors to:

assign each condition entity a score based on features of a condition represented by the condition entity; and

rank the condition entities based on respective scores.

10. The system of claim 9 , wherein the instructions to rank the condition entities further cause the one or more processors to:

for each condition entity:

score evidence entities corresponding to each of the condition entities based on properties of the evidence entities;

rank the evidence entities based on respective scores; and

select a top N scoring evidence entities as most indicative evidence for the condition entity.

11. The system of claim 1 , wherein the instructions to link entities further cause the one or more processors to:

invalidate a subset of links between evidence entities and condition entities in the domain-specific ontologies wherein the subset of the links are deleted.

12. The system of claim 1 , wherein the electronic document is an unstructured document.

13. A computer-implemented method of processing an electronic document comprising:

receiving the electronic document including component documents,

the component documents being produced from different data sources;

extracting plain text data of the component documents;

obtaining document structure information and styling information of the electronic document from the component documents;

automatically segmenting into sections, the plain text data,

the automatically segmenting performed by analyzing contents of the component documents using boundary rules, and

the boundary rules specifying grouping constraints on the plain text data;

extracting entities automatically from the plain text data using machine learning (ML), natural language processing (NLP) and entity recognition (ER);

categorizing the entities into one of condition entities or evidence entities that support the condition entities based on an entity type;

linking the supporting evidence entities to the condition entities within the plain text data;

confirming accuracy of a condition presented in the electronic document if a score of one of the condition entities associated with the condition is a highest score among scores of the condition entities present in the electronic document; and

causing to display on an interactive GUI, another of the condition entities with the highest score among the scores of the condition entities as an indicator of an accurate condition if the score of the condition entity associated with the condition is not highest among scores of the condition entities present in the electronic document.

14. The method of claim 13 , further comprising:

scoring the condition entities using domain-specific ontologies and inferences from knowledge bases; and

ranking the condition entities based on the scores of the condition entities.

15. The method of claim 13 wherein extracting entities automatically from the plain text data using ML further comprises:

identifying dates in the plain text data; and

determining boundaries between the sections of the plain text data based at least on the dates.

16. The method of claim 13 wherein the electronic document is an unstructured document, and extracting the plain text data from the component files further comprises:

generating a markup file from the unstructured document,

the markup file including information from the unstructured document, and

the information being styled and formatted as presented in the component documents; and

parsing the markup file for the extraction of the plain text data.

17. The method of claim 13 wherein the electronic document is a structured document.

18. The method of claim 13 , further comprising:

enabling user interactions for editing boundaries between the sections, the condition entities, the evidences entities and relations therebetween;

transmitting user edits to one or more of content, structure and format to knowledge bases; and

enabling updates to knowledge graphs within the knowledge bases so that future queries to the knowledge bases receive responses that include the user edits.

19. A non-transitory storage medium comprising machine-readable instructions that cause at least one processor to:

convert a non-editable image file pertaining to an electronic document including at least one component document into a markup file, wherein the markup file preserves formatting and structure of the component document from the image file;

parse the markup file to extract plain text data of the image file and styling information pertaining to the formatting and document structure information of the image file;

automatically segment into sections the plain text data, by analyzing contents of the markup file using at least boundary rules;

identify and extract entities automatically from the plain text data using natural language processing (NLP);

link the entities within the plain text using domain-specific ontologies, knowledge bases and graphical inferences;

generate an interactive GUI that displays content including the plain text data formatted according to the styling information,

the GUI including annotated entity relations derived from the electronic document, and

the GUI enabling user interactions for editing the boundaries, condition entities and evidences entities and relations therebetween; and

transmit user edits to one or more of the content, structure and format to the knowledge bases.

20. The non-transitory storage medium of claim 19 , wherein the instructions for linking entities further comprise instructions to:

categorize the entities into one of condition entities or evidence entities based on an entity type; and

link each of the condition entities to one or more of the evidence entities within the plain text using domain-specific ontologies, knowledge bases and graphical inferences.