US 11,960,849 B2Granted

Using machine learning models to analyze contractual terms and clauses in a legal contract, to recommend edits, and to make changes to a workflow

AssigneeIronclad, Inc.

InventorsCai GoGwilt, Jennifer S. S. Monteleone, Adam Weber, Yujiao Zhang, Angela Kou, Vidya Ravikumar, Kevin Verdieck, Wolfgang Van HellicksonSabelhaus, Katherine Vilhena, Peter Nam That Ton, Nilay Amit Sadavarte, Sumuk Rao, Jean-Marc Soumet, Alexander S. Gillmor

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

Aspects discussed herein may relate to using machine learning models as part of methods and techniques for ingesting, creating, storing, editing, and managing a document. The document may be a legal contract that includes one or more clauses. Among other things, one or more machine learning models may be configured to recognize clauses and/or classifications, or types, of clauses. For example, the one or more generative language models may be used to generate one or more recommended edits to a clause, generate one or more suggested clauses that are missing from the contract, and/or generate one or more suggested locations where a clause may be inserted into or moved within the contract.

Claims20

Dependent claims

A method comprising:

receiving, by a computing platform of a networked document workflow environment that enables a document to be edited by a user, first user input that indicates at least a portion of the document that is to be analyzed by one or more machine learning models, wherein the portion of the document includes a first clause for a legal contract;
based on the first user input, causing at least the portion of the document to be analyzed by the one or more machine learning models by:
from a plurality of natural language processing models configured to process contractual terms of the legal contract, process clauses of the legal contract, and determine suggested edits to the legal contract, selecting, by the computing platform, two or more natural language processing models that will process the portion of the document, wherein the two or more natural language processing models includes an extractive language model that is configured to classify clauses of the legal contract as one or more types of clauses and a generative language model that is configured to determine suggested edits to the one or more types of clauses,
using, by the computing platform and based on the selecting, the two or more natural language processing models to process the portion of the document, resulting in output data from the two or more natural language processing models, wherein the output data includes a classification of the portion of the document as a first type of the one or more types of clauses, and wherein the output data includes, based on the first type, a first suggested edit to the portion of the document, and
determining, by the computing platform, response data based on the output data, wherein the response data indicates the first suggested edit to the portion of the document;
and
causing, by the computing device, display of the response data to provide a visual indication of the first suggested edit to the user in the networked document workflow environment.

4 more elements

A method comprising:

receiving, by a computing platform of a networked document workflow environment that enables a document to be edited by a user based on configured rules, first user input that indicates one or more edits to at least a portion of the document, wherein the portion of the document includes a clause for a legal contract, and wherein the configured rules indicate user-configured language for the clause;
based on the first user input, causing at least the portion of the document to be analyzed by one or more machine learning models by:
from a plurality of natural language processing models configured to process contractual terms of the legal contract, process clauses of the legal contract, and determine suggested edits to the legal contract, selecting, by the computing platform, two or more natural language processing models that will process the portion of the document, wherein the two or more natural language processing models includes an extractive language model that is configured to classify clauses of the legal contract as one or more types of clauses and a generative language model that is configured to determine suggested edits to the one or more types of clauses based on the user-configured language,
using, by the computing platform and based on the selecting, the two or more natural language processing models to process the portion of the document, resulting in output data from the two or more natural language processing models, wherein the output data includes a classification of the portion of the document as a first type of the one or more types of clauses, wherein the output data includes, based on the first type, a first suggested edit to the portion of the document, and wherein the first suggested edit is based on the user-configured language, and
determining, by the computing platform, response data based on the output data, wherein the response data indicates the first suggested edit to the portion of the document;
and
causing, by the computing device, display of the response data to provide a visual indication of the first suggested edit to the user in the networked document workflow environment.

4 more elements

A method comprising:

receiving, by a computing platform of a networked document workflow environment that enables a document to be edited by a user based on configured rules, first user input that indicates at least a first portion of the document that is to be analyzed by one or more machine learning models, wherein the first portion of the document includes a first clause for a legal contract;
based on the first user input, causing at least the first portion of the document to be analyzed by:
from a plurality of natural language processing models configured to process contractual terms of the legal contract, process clauses of the legal contract, and determine suggested edits to the legal contract, selecting, by the computing platform, a first set of natural language processing models that will process the first portion of the document, wherein the first set of natural language processing models includes an extractive language model that is configured to classify clauses of the legal contract as one or more types of clauses and a generative language model that is configured to determine suggested edits to the one or more types of clauses,
using, by the computing platform and based on the selecting, the first set of natural language processing models to process the first portion of the document, resulting in first output data from the first set of natural language processing models, wherein the first output data includes a classification of the first portion of the document as a first type of the one or more types of clauses, and wherein the first output data includes, based on the first type, a first suggested edit to the first portion of the document, and
determining, by the computing platform, first response data based on the first output data, wherein the first response data indicates the first suggested edit to the first portion of the document;
causing, by the computing device, display of the first response data to provide a visual indication of the first suggested edit to the user in the networked document workflow environment;
receiving, by the computing platform, second user input that indicates one or more edits to at least a second portion of the document, wherein the second portion of the document includes a second clause for a legal contract, and wherein the configured rules indicate user-configured language for the second clause;
based on the second user input, causing at least the second portion of the document to be analyzed by:
selecting, by the computing platform and from the plurality of natural language processing models, a second set of natural language processing models that will process the second portion of the document, wherein the second set of natural language processing models includes the extractive language model and the generative language model,
using, by the computing platform and based on the selecting, the second set of natural language processing models to process the second portion of the document, resulting in second output data from the second set of natural language processing models, wherein the second output data includes a classification of the second portion of the document as a second type of the one or more types of clauses, wherein the second output data includes, based on the second type, a second suggested edit to the second portion of the document, and wherein the second suggested edit is based on the user-configured language, and
determining, by the computing platform, second response data based on the second output data, wherein the second response data indicates the second suggested edit to the second portion of the document;
and
causing, by the computing device, display of the second response data to provide a visual indication of the second suggested edit to the user in the networked document workflow environment.

10 more elements

Dependencies are inferred from the claim text.

Facts

Application: 18/467,363Filed 14 Sep 2023 · us provisional application
Priority: 14 Sep 2022
Published: US 2024/0086651 A114 Mar 2024
Granted: 16 Apr 2024
Term ends: Not estimatedThe source records don’t support an estimate.

Classification4 codes

G06F 40/56
Rule-based translation: natural language generation
G06F 40/131
Use of codes for handling textual entities: fragmentation of text files, e.g. creating reusable text-blocks; Linking to fragments, e.g. using…
G06F 40/166
Text processing: editing, e.g. inserting or deleting
G06F
Computing and data processing