Context-Aware Sensitive Information Redaction Using Entity Identifiers

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing document redaction tools lack context awareness, leading to inefficient and resource-intensive manual corrections and inappropriate redaction of sensitive information.

Innovation Solution

Implement a context-aware redaction process that filters sensitive items based on entity identifiers, decoupling context-aware filtering from sensitive information detection to enhance efficiency and scalability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If automated sensitive information detection is applied without context awareness, then detection coverage is improved, but inappropriate redaction increases and manual correction effort increases

Engineering Contradiction:

Improveautomated sensitive information detectionVSAvoidappropriateness of redaction

Core Design Contradiction:

Extent of automationVSReliability

Solution Approach 1:

The patent introduces context information (entity identifiers, document metadata, user profiles) as an intermediary layer between sensitive information detection and redaction decisions. This intermediary contextualizes detected sensitive items, enabling the system to distinguish between information that should and should not be redacted, thereby maintaining high automation while improving reliability.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system dynamically adjusts redaction parameters based on context. Instead of applying fixed rediction rules, the system modifies detection sensitivity, redaction scope, and confidence thresholds according to contextual parameters such as document type, entity relationships, and user permissions, resolving the contradiction between comprehensive detection and appropriate redaction.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If context-aware filtering is implemented, then redaction accuracy is improved, but system complexity increases

Engineering Contradiction:

Improveredaction accuracyVSAvoidsystem architecture complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the redaction system into distinct modular components: sensitive information detector, context analyzer, decision engine, and redaction executor. Each module handles a specific aspect of the task, improving accuracy through specialized processing while managing complexity through clear separation of concerns and independent module development.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The context-aware filtering layer is designed as a universal component that can be applied across different document types, sensitive information categories, and organizational policies. This multi-functional design achieves high redaction accuracy across diverse scenarios without proportionally increasing system complexity, as the same core mechanisms serve multiple purposes.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Quantity of substance

If comprehensive sensitive information detection is performed, then detection completeness is improved, but computational resource consumption increases

Engineering Contradiction:

Improvedetection completenessVSAvoidcomputational resource consumption

Core Design Contradiction:

Quantity of substanceVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary filtering using context information before applying comprehensive detection algorithms. By pre-identifying document characteristics, entity types, and relevant sensitivity categories, the system narrows the detection scope in advance, achieving comprehensive detection of relevant sensitive information while avoiding unnecessary computational expenditure on irrelevant content.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent applies detection intensity proportional to contextually determined needs. Instead of uniformly applying maximum detection thoroughness to all documents, the system adjusts detection depth and breadth based on contextual risk assessment, achieving sufficient detection completeness for each document type while optimizing computational resource allocation across the document population.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12367250B2Detection and removal of predefined sensitive information types from electronic documents

Publication Date: 2025.07.22 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Automated and semi-automated document redaction technology is disclosed herein. In certain example embodiments, ‘context-aware’ redaction is provided. Automated techniques are used to identify a set of potentially sensitive item(s) within a document. The potentially sensitive item(s) are filtered based on contextual information, such an entity identifier (e.g. person identifier, person group identifier identifying a group of multiple people, organization identifier etc.), resulting in a filtered set of redaction candidate(s). The filtered redaction candidate(s) may, for example, be redacted from the document automatically, or outputted as suggestions in an assisted redaction tool, e.g. via a document redaction graphical user interface. Other example embodiments consider selective redaction when uploading and/or downloading documents via a proxy server, to prevent intended or unintended release of potentially sensitive information, e.g. in a web browsing context.