Document Risk Analysis via Knowledge Graph Visualization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Users may make decisions based on unreliable information in documents, leading to potential harm due to the variability in quality and reliability of information across different sources.

Innovation Solution

A method that analyzes documents to identify references, generates a knowledge graph to represent information, and applies a visual indicator to emphasize potential problems based on a risk assessment, using natural language processing and graph theory to determine reliability and integrity across references.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If documents from various sources are made readily available to users, then information accessibility and quantity are improved, but information reliability and consistency deteriorate

Engineering Contradiction:
Improveinformation accessibilityVSAvoidinformation reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent introduces an intermediary system that acts as a mediator between users and documents. This system automatically analyzes documents, generates knowledge graphs, assesses risks, and applies visual indicators to highlight problematic information. The intermediary process resolves the contradiction by maintaining broad document accessibility while filtering and marking unreliable information, allowing users to access diverse sources without being misled by inconsistent or unreliable content.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If users are provided with unrestricted access to information across multiple sources, then information variety and completeness are improved, but the risk of harmful decisions increases

Engineering Contradiction:
Improveinformation varietyVSAvoidrisk of harm
Core Design Contradiction:
Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The patent implements preliminary action by automatically analyzing documents and applying risk assessments before users consume the information. The system proactively identifies potentially harmful or unreliable content, generates knowledge graphs to verify consistency, and applies visual indicators to warn users in advance. This preliminary filtering and marking process allows users to access diverse information sources while being protected from harmful content through advance warning signals.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If comprehensive document analysis is performed to ensure information reliability, then information quality is improved, but processing time and computational resources increase

Engineering Contradiction:
Improveinformation qualityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies local quality by focusing analysis efforts on specific critical elements within documents rather than uniformly analyzing entire documents. The system identifies and prioritizes key references, claims, and data points that most impact reliability assessment. By concentrating computational resources on locally critical sections and using visual indicators to highlight only problematic areas, the system maintains high information quality assessment while reducing overall processing time compared to comprehensive full-document analysis.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS11423094B2Document risk analysis
Publication Date: 2022.08.23 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US11423094B2 patent drawing
  • US11423094B2 patent drawing
  • US11423094B2 patent drawing

AI summary

A system and method for assessing a potential problem associated with information in a document. In an embodiment, a document for analysis may be received, the document being part of a corpus of one or more documents stored in an electronic format. The document may be analyzed to identify reference(s) in the document, wherein the reference(s) are each selected from the group comprising a natural language statement, a reference in a first part of the document to a second part of the document, or a reference to or from another document. Based on the reference(s), a knowledge graph may be generated to represent information in the document. A risk assessment of a reference in the document may be determined using the knowledge graph. A visual indicator may be applied to the document that emphasizes a potential problem with the reference based on the risk assessment.