US 11,455,464 B2Granted

Document content classification and alteration

AssigneeAccenture Global Solutions Limited

InventorsIan Hook, Jose Eduardo Arias Quiros, Rebeca Gomez Vez, Marko Krema, Thomas A. Hsu

Life of this patent

Estimating the term…

Loading patent…

Abstract

A device may receive a document including text, images, and one or more embedded documents. The device may parse the document to identify a content segment in the document. The device may determine a context for the content segment, wherein the context includes at least one of an entity identified by the content segment, a semantic meaning of the content segment, or an object identified by the content segment. The device may classify the content segment using a content classification model and based on the context for the content segment. The device may selectively alter the content segment based at least in part on a set of alteration rules, to generate a modified document. The device may provide the modified document based on selectively altering the content segment.

Claims20

Dependent claims

A method, comprising:

receiving, by a device, a document including at least one of text, images, and one or more embedded documents;
parsing, by the device, the document to identify a content segment in the document,
wherein the content segment is at least one of a text segment or an image, from at least one of the images of the document or an embedded document of the one or more embedded documents of the document, and
wherein the content segment has multiple possible semantic meanings based on different types of context;
determining, by the device, a context for the content segment,
wherein the context is determined based on using natural language processing of other words that are outside the content segment in the document to determine a semantic meaning, from the multiple possible semantic meanings, of the content segment;
classifying, by the device, the content segment using a content classification model and based on the context for the content segment,
wherein the content segment is classified into a first type of content segment that is proposed to be altered or a second type of content segment that is proposed to remain unaltered;
determining, by the device and based on classifying the content segment into the first type of content segment, a suggested replacement for the first type of content segment;
providing, by the device and for display via a user interface, information identifying the content segment, the first type of content segment, and the suggested replacement;
receiving, by the device and via the user interface, a set of selections associated with the information identifying the content segment, the first type of content segment, and the suggested replacement;
selectively altering, by the device, the content segment based at least in part on a set of alteration rules, the semantic meaning, and the set of selections, to generate a modified document,
wherein the content segment is altered to a generic version of the content segment to preserve a threshold quantity of the semantic meaning;
providing, by the device, the modified document based on selectively altering the content segment;
updating, by the device, the content classification model based on the set of selections;
and
storing, by the device, the updated content classification model.

14 more elements

A device, comprising:

one or more memories;
and
one or more processors communicatively coupled to the one or more memories, the one or more processors configured to:
receive a document;
parse the document to identify a set of content segments in the document;
determine a set of contexts for the set of content segments,
wherein the set of contexts is determined based on using natural language processing of other words that are outside the set of content segments in the document to determine a respective set of semantic meanings of the set of content segments;
classify the set of content segments using a content classification model and based on the set of contexts for the set of content segments,
wherein a first subset of the set of content segments is classified into a type of content segment that is proposed to be altered, and a second subset of the set of content segments is classified into a type of content segment that is proposed to remain unaltered;
determine a suggested replacement based on the respective set of semantic meanings for the type of content segment;
provide, for display via a user interface, information identifying the first subset of the set of content segments, the second subset of the set of content segments, and the suggested replacement;
receive, via the user interface, a set of selections associated with the displayed information identifying the first subset of the set of content segments and the suggested replacement;
alter the first subset of the set of content segments based at least in part on the suggested replacement to generate a modified document,
wherein the one or more processors, when selectively altering the content segment, are to:
alter the content segment based on the set of selections, or
retain the content segment based on the set of selections;
and
provide the modified document based on altering the first subset of the set of content segments.

15 more elements

A non-transitory computer-readable medium storing instructions, the instructions comprising:

one or more instructions that, when executed by one or more processors, cause the one or more processors to:
obtain a training data set including a plurality of documents;
train, using the training data set and a machine learning technique, a content classification model;
store the content classification model for use in classifying a document;
receive, after storing the content classification model, the document;
parse the document to identify a set of content segments in the document;
determine a respective set of contexts for the set of content segments,
wherein the respective set of contexts is determined based on using natural language processing of other words that are outside the set of content segments in the document to determine a respective set of semantic meanings of the set of content segments;
classify the set of content segments using the stored content classification model,
wherein a first subset of the set of content segments is classified into a first type of content segment that is proposed to be altered and a second subset of the set of content segments is classified into a second type of content segment that is proposed to remain unaltered;
alter the first subset of the set of content segments, based on classifying the set of content segments, to generate a modified document;
provide the modified document based on altering the first subset of the set of content segments;
update the stored content classification model based on one or more alterations to the first subset of the set of content segments;
and
store the updated content classification model.

12 more elements

Dependencies are inferred from the claim text.