N-gram Dictionary Data Obfuscation for Privacy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data protection methods fail to effectively obfuscate sensitive information while maintaining the statistical properties necessary for research and analysis, particularly in large datasets like medical records, without exposing confidential data.

Innovation Solution

A data obfuscation system using n-grams, where a dictionary module maps n-grams to dictionary terms, allowing for secure data analysis while maintaining statistical properties, and an obfuscation module generates tokens for further research, employing machine learning algorithms on encrypted data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional data protection methods are used to obfuscate sensitive information, then confidentiality is improved, but statistical properties necessary for research and analysis are lost

Engineering Contradiction:

ImproveconfidentialityVSAvoidstatistical properties

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The patent introduces n-grams as an intermediary representation that bridges the gap between raw sensitive data and obfuscated data. By converting data into n-gram sequences and applying transformations only to these intermediate representations, the system preserves statistical properties while achieving obfuscation. The n-grams act as a mediator that maintains analytical utility without exposing sensitive information.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent applies parameter changes by transforming the data representation from raw values to n-gram sequences, and then further transforming these n-grams through various obfuscation techniques. This changes the parameters of the data while maintaining its statistical structure, allowing research and analysis to proceed on obfuscated data without losing important statistical properties.

Inventive Principle:

Principle #35Parameter changes

2Reliability

If data is obfuscated to protect privacy, then confidentiality is improved, but the ability to perform pattern identification and predictive analysis deteriorates

Engineering Contradiction:

ImproveconfidentialityVSAvoidresearch and analysis capability

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent creates a copied representation of the data in the form of n-grams that preserves the statistical structure and patterns needed for research and analysis. This copy can be analyzed and transformed without affecting the original sensitive data, maintaining both confidentiality and analytical capability simultaneously.

Inventive Principle:

Principle #26Copying

3Reliability

If n-grams are transformed to obfuscate data, then confidentiality is improved, but data structure and meaning are lost

Engineering Contradiction:

ImproveconfidentialityVSAvoiddata structure

Core Design Contradiction:

ReliabilityVSLoss of information

Solution Approach 1:

The patent segments data into n-gram units, which preserves the structural relationships and contextual meaning while enabling effective obfuscation. By working with segmented n-grams rather than raw data or completely anonymized aggregates, the system maintains data structure and meaning necessary for meaningful analysis.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS11568080B2Systems and method for obfuscating data using dictionary

Publication Date: 2023.01.31 SOLVENTUM INTELLECTUAL PROPERTIES CO

AI summary

At least some aspects of the present disclosure feature systems and methods for obfuscating data. The method includes the steps of receiving an input data stream including a sequence of n-grams, mapping at least some of the sequence of n-grams to corresponding dictionary terms using a dictionary, and disposing the corresponding tokens to an output data stream.