US 10,482,540 B2Granted

Data translation

AssigneeAccenture Global Solutions Limited

InventorsChung-Sheng Li, Guanglei Xiong, Sukryool Kang, Ashish Jain, Colin Connors, Benjamin Nathan Grosof, Neeru Narang

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

A classifier receives policy data corresponding to a new policy. Further, the classifier processes the policy data to classify the policy data into an obligation class and an informational class. An information extractor then extracts metadata from the policy data that is classified into the obligation class. Subsequently, a data translator determines if there is an incremental change in the policy data based on a comparison of the policy data with policy data corresponding to existing policies. On determining the incremental change in the policy data, the data translator translates the policy data that is classified into the obligation class into a rule based on the metadata. A rules engine then receives the rule from the data translator for claims adjudication.

Claims16

Dependent claims

A system to translate of policies into rules for claims adjudication, the system comprising:

a classifier to:
receive policy data corresponding to a new policy, the policy data comprising natural language content to be selectively translated to a new rule, wherein the new rule comprising a procedure to determine whether a claim complies with pre-determined requirements pertaining to claim adjudication;
and
process a portion of the policy data to classify the portion of the policy data into an obligation class and an informational class, wherein the obligation class comprises useful information to be used for generating the new rule and the informational class refers to the class that signifies additional information or non-useful data that would not be used for translation of the new policy into rules;
an information extractor, in combination with the classifier, the information extractor to extract metadata from the portion of the policy data that is classified into the obligation class;
a data translator, in combination with the information extractor, the data translator to:
determine whether there is an incremental change in the portion of policy data based on a comparison of the portion of the policy data with existing policy data corresponding to existing policies, wherein the existing policy data comprises knowledge graphs of the existing policies and a plurality of rules corresponding to the existing policies, wherein the existing policies are decomposed and curated into multiple knowledge representations to form the knowledge graphs, and the comparison is based on at least one of topic classification, entities and relationship, and knowledge representations extracted from the new policy vs. the existing policies;
and
on determining the incremental change in the portion of the policy data, selecting the portion of the policy data that is classified into the obligation class and discarding the portion of the policy data that is classified into the information class;
and
translate the selected portion of policy data that is classified into the obligation class into the new rule based on the metadata;
and
a rules engine, in combination with the data translator, the rules engine to receive the new rule from the data translator for the claim adjudication based in part on the portion of the new policy.

10 more elements

A system to translate of policies into rules for claims adjudication, the system comprising:

a classifier to:
receive policy data corresponding to a new policy, the policy data comprising natural language content to be selectively translated to a new rule, wherein the new rule comprising a procedure to determine whether a claim complies with pre-determined requirements pertaining to claim adjudication;
and
process a portion of the policy data to classify the portion of the policy data into an obligation class and an informational class, wherein the obligation class comprises useful information to be used for generating the new rule and the informational class refers to the class that signifies additional information or non-useful data that would not be used for translation of the new policy into rules;
an information extractor, in combination with the classifier, the information extractor to extract metadata from the portion of the policy data that is classified into the obligation class;
a data translator, in combination with the information extractor, the data translator to:
determine whether there is an incremental change in the portion of policy data based on a comparison of the portion of the policy data with existing policy data corresponding to existing policies, wherein the existing policy data comprises knowledge graphs of the existing policies and a plurality of rules corresponding to the existing policies, wherein the existing policies are decomposed and curated into multiple knowledge representations to form the knowledge graphs, and the comparison is based on at least one of topic classification, entities and relationship, and knowledge representations extracted from the new policy vs. the existing policies;
and
on determining the incremental change in the portion of the policy data, selecting the portion of the policy data that is classified into the obligation class and discarding the portion of the policy data that is classified into the information class;
and
translate the selected portion of policy data that is classified into the obligation class into the new rule based on the metadata.

8 more elements

A computer-implemented method to translate of policies into rules for claims adjudication, executed by at least one processor, the method comprising:

receiving policy data corresponding to a new policy, the policy data comprising natural language content to be selectively translated to a new rule, wherein the new rule comprising a procedure to determine whether a claim complies with pre-determined requirements pertaining to claim adjudication;
processing a portion of the policy data to classify the portion of the policy data into an obligation class and an informational class, wherein the obligation class comprises useful information to be used for generating the new rule and the informational class refers to the class that signifies additional information or non-useful data that would not be used for translation of the new policy into rules;
extracting metadata from the portion of the policy data that is classified into the obligation class;
determining whether there is an incremental change in the portion of policy data based on a comparison of the portion of the policy data with existing policy data corresponding to existing policies, wherein the existing policy data comprises knowledge graphs of the existing policies and a plurality of rules corresponding to the existing policies, wherein the existing policies are decomposed and curated into multiple knowledge representations to form the knowledge graphs, and the comparison is based on at least one of topic classification, entities and relationship, and knowledge representations extracted from the new policy vs. the existing policies;
and
on determining the incremental change in the portion of the policy data, selecting the portion of the policy data that is classified into the obligation class and discarding the portion of the policy data that is classified into the information class;
and
translate the selected portion of policy data that is classified into the obligation class into the new rule based on the metadata.

5 more elements

Dependencies are inferred from the claim text.

Facts

Application: 15/887,813Filed 2 Feb 2018
Published: US 2019/0244300 A18 Aug 2019
Granted: 19 Nov 2019
Term ends: Not estimatedThe source records don’t support an estimate.

Classification9 codes

G06Q 40/08
Finance; Insurance; Tax strategies; Processing of corporate or income taxes: insurance
G06F 16/285
Relational databases: clustering or classification
G06N 5/04
Computing arrangements using knowledge-based models: inference or reasoning models
G06N 20/00
Machine learning

Family

US 2019/0244300 A1Pre-grant publication · published 8 Aug 2019
US 10,482,540 B2This patent

From the US filing history. Foreign family members aren’t listed.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key