Font Information Embedding PII Indicators for Document Security
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for identifying and managing personally identifiable information (PII) within documents are inefficient, requiring multiple models and high processing and memory resources, leading to overburdened systems and inaccurate determinations.
Innovation Solution
The system uses font information, such as vector images, pixel matrices, or stroke paths, to embed PII indicators within documents, allowing for efficient identification and decisioning on PII, with rules engines determining access rights and redacting sensitive information as needed, and storing redundant PII only once across documents.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used to identify and manage PII within documents, then accurate determination can be achieved, but processing and memory requirements become excessively high
Solution Approach 1:
The patent combines multiple PII identification models into a single unified model that can handle different PII types (personal names, addresses, phone numbers, etc.) simultaneously. This consolidation reduces memory requirements by eliminating the need to load and manage multiple separate models, while maintaining accurate PII identification across various document types and formats.
Solution Approach 2:
The patent develops a universal PII identification system that can process multiple types of personally identifiable information (names, addresses, phone numbers, email addresses, social security numbers, etc.) through a single model. This multi-functional approach reduces memory consumption compared to maintaining separate specialized models for each PII type, while preserving accurate identification capabilities across all PII categories.
2Adaptability or versatility
If multiple models are used for different PII entities, then comprehensive identification can be achieved, but device complexity increases
Solution Approach 1:
The patent merges multiple PII identification models into a single unified model that can identify various types of personally identifiable information (names, addresses, phone numbers, email addresses, social security numbers, etc.). This consolidation reduces system complexity by eliminating the need to manage multiple separate models, their configurations, and interconnections, while maintaining comprehensive PII type coverage through the unified model's multi-classification capability.
3Measurement precision
If traditional PII identification systems are implemented, then PII can be detected, but processing requirements become excessively high
Solution Approach 1:
The patent combines multiple PII identification models into a single unified model, reducing processing requirements by eliminating redundant computations that would occur if multiple separate models were executed. The unified model processes document text through a single pipeline, maintaining accurate PII detection capability while significantly reducing CPU usage, memory access overhead, and overall processing time compared to traditional multi-model approaches.
4Stability of the object's composition
If redundant PII text is stored in each document, then document integrity is maintained, but storage requirements increase
Solution Approach 1:
The patent implements a reference-based storage system where redundant PII text is stored only once in a centralized repository, and multiple documents that contain the same PII reference this single copy through pointers or references. This approach maintains document integrity by preserving references to the original PII data while dramatically reducing storage requirements, as duplicate PII content is eliminated and only unique PII instances are fully stored.
Data Source
AI summary
Information determination and decisioning systems that allow for the classification, identification, and decisioning of personally identifiable information (hereinafter “PII”) that is located within documents. Text within a document may be identified as PII text and/or non-PII text, and PII indicators may be associated with the font of the text in order to define and be able to track the PII text within the document. The PII indicators may provide different information about the PII text, such as indicating that the text includes PII, the PII type associated with the text, the locations in which the text constitutes PII. The PII indicators may be stored within the font information of the font, such as through data stored by the font and/or the data used to create the font (e.g., vector images that define the font curves, the dot matrices that define the font, the stroke paths that define the font, etc.).