Unstructured Data Entity Extraction for Contract Generation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Small business owners face inefficiencies and errors when creating invoices and contracts from unstructured email data, as manually copying information is time-consuming and prone to errors.

Innovation Solution

A system and method that extracts entities from unstructured data sources using natural language processing, generates a contract model, and predicts contract details by correlating structured data, providing a proposed contract with confidence scores that can be edited and refined by users.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual copying of information from email threads into invoices is performed, then the invoice can be created, but the process is time-consuming and error-prone

Engineering Contradiction:
Improveinvoice creation speedVSAvoidtime spent on manual copying
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent replaces the mechanical manual copying process with an automated natural language processing system that extracts entities from unstructured email data and populates invoice fields automatically, eliminating the need for manual transcription while maintaining accuracy

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables the invoice creation process to serve itself by automatically extracting relevant information from email threads and populating the invoice template without human intervention, allowing the process to complete autonomously

Inventive Principle:
Principle #25Self-service

2Reliability

If manual copying of information from email threads into invoices is performed, then the invoice can be created, but errors may occur in the invoice

Engineering Contradiction:
Improveinvoice accuracyVSAvoidcomplexity of data extraction process
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent replaces error-prone manual copying with an automated NLP-based entity extraction system that systematically identifies and extracts relevant information from unstructured email data, reducing human error while managing complexity through algorithmic processing

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces an intermediary natural language processing layer between the unstructured email data and the invoice template, which acts as a mediator to extract, validate, and map entities appropriately, ensuring accuracy while handling the complexity of data transformation

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If similar issues arise when sending contracts based on unstructured data sources, then the contract can be generated, but the process remains inefficient and error-prone

Engineering Contradiction:
Improvecontract generation efficiencyVSAvoidease of contract creation
Core Design Contradiction:
ProductivityVSEase of operation

Solution Approach 1:

The patent creates a universal entity extraction system that can handle multiple document types (invoices, contracts, agreements) from unstructured data sources, making the system multi-functional and applicable across different business documentation needs without requiring separate manual processes for each document type

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS10733240B1Predicting contract details using an unstructured data source
Publication Date: 2020.08.04 INTUIT INC
  • US10733240B1 patent drawing
  • US10733240B1 patent drawing
  • US10733240B1 patent drawing

AI summary

A method may include extracting first entities from a first portion of an unstructured data source associated with a user, obtaining, based on the first entities, a contract model including elements and a contract type, generating, by applying the contract model to the first entities, a proposed contract including a contract score and, for each element, element values. Each element value may include an element value score. The method may further include identifying a structured data source associated with the user, obtaining, from the structured data source, structured data corresponding to the first entities, correlating the structured data with an element value of the proposed contract, and modifying, by applying the contract model to the structured data, the element value score of the element value.