Document Graph Structure for Semantic Agreement Retrieval

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional document management systems lack the ability to support complex queries related to the location of electronic documents and the generation of documents using an existing corpus, struggle with tracking entities involved in agreement negotiations, and fail to determine how agreements have been amended over time, leading to incomplete or incorrect information retrieval.

Innovation Solution

A document graph structure is implemented, utilizing machine learning models to transform unstructured text into structured nodes and edges, enabling semantic similarity searches to accurately retrieve relevant information within electronic documents.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional document management systems are used to query agreement storage databases, then basic document retrieval is possible, but the queries return incomplete data or incorrect information

Engineering Contradiction:

Improvesearch accuracyVSAvoidcompleteness of retrieved data

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent replaces conventional keyword-based mechanical search systems with an AI-powered semantic search system that uses natural language processing and machine learning to understand the meaning and context of queries, enabling accurate retrieval of complete agreement information including parties, clauses, and amendments

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent introduces an AI intermediary layer between the user query and the document database that processes, understands, and translates natural language queries into comprehensive document retrieval operations, ensuring complete and accurate information retrieval

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If existing agreements are stored for future reference, then prior agreements can be helpful for generation of future agreements, but the existing agreements need to be tailored to specific contractual obligations which is extremely difficult

Engineering Contradiction:

Improvereusability of prior agreementsVSAvoiddifficulty of tailoring agreements

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments agreements into standardized modular components such as parties, clauses, terms, and conditions using AI-powered document structure analysis. This segmentation enables easy recombination and tailoring of existing agreement components for new agreements, reducing the complexity of adaptation while maintaining reusability

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent uses AI to automatically identify and change parameters such as party names, dates, and specific terms when adapting existing agreements to new contractual obligations, making the tailoring process systematic and manageable rather than extremely difficult

Inventive Principle:

Principle #35Parameter changes

3Quantity of substance

If document management systems store large collections of electronic documents, then comprehensive agreement history is available, but tracking entities involved in negotiations and amendments over time becomes unmanageable

Engineering Contradiction:

Improvevolume of document collectionVSAvoidcomplexity of tracking entities and amendments

Core Design Contradiction:

Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent implements a universal AI-powered entity recognition and relationship tracking system that automatically identifies and tracks all entities (parties, representatives, negotiating entities) and their relationships across the entire document collection, managing complexity through centralized intelligent processing rather than separate tracking mechanisms

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS20260050638A1Document graph

Publication Date: 2026.02.19 DOCUSIGN INC

AI summary

A method, an apparatus, and a computer-readable storage medium for generating a document graph. A plurality of electronic documents is received. Each electronic document has a predetermined document type. A machine learning model is selected from the plurality of machine learning models based on the predetermined document type. The selected machine learning model is instructed to extract a plurality of document portions from each electronic document in the plurality of electronic documents in accordance with the predetermined document type. A relationship between two or more document portions is defined based on a content of each document portion, and the document portions are associated based on the relationship. A graph structure having a plurality of nodes is generated. Each node includes at least one document portion. Each node is connected to another node in accordance with the relationship between document portions included in the nodes. The graph structure is stored.