Anonymous Biometric Trajectories Using Locality-Sensitive Hashing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric systems struggle to track and study human behavior over time without violating individuals' privacy rights, as they often require identifying and storing specific biometric data, which is legally and ethically prohibited.
Innovation Solution
A system and method using one-way locality-sensitive hashing to convert biometric data into anonymous biometric keys, allowing the creation of biometric trajectories that anonymize data from multiple individuals, enabling the tracking of behavioral patterns over time without identifying specific subjects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If biometric data is stored and tracked to study behavior patterns over time, then behavioral analysis capability is improved, but privacy protection deteriorates
Solution Approach 1:
The patent extracts only the essential features needed for behavioral analysis while removing personally identifiable information. Biometric data is processed through hashing functions that extract behavioral patterns without retaining information that could identify specific individuals, thus separating useful analytical data from privacy-sensitive information.
Solution Approach 2:
The patent introduces hash functions and anonymization algorithms as intermediary processing steps between data collection and storage. These intermediaries transform raw biometric data into anonymous representations that preserve behavioral information while eliminating personal identifiers, acting as a buffer that protects privacy while enabling analysis.
2Measurement precision
If specific biometric data is stored to identify individuals, then tracking precision is improved, but anonymity deteriorates
Solution Approach 1:
The patent applies different processing qualities to different aspects of biometric data. Certain features are processed with high precision for behavioral pattern recognition, while other features are deliberately degraded or removed to prevent identification. This local differentiation allows simultaneous achievement of tracking precision and anonymity.
Solution Approach 2:
The patent transforms biometric data parameters through mathematical hashing functions that change the data representation. The transformation preserves essential behavioral parameters while altering or eliminating identification parameters, enabling the system to maintain tracking precision for behavior analysis while losing the ability to identify specific individuals.
3Duration of action of moving object
If biometric trajectories are stored for long-term behavior study, then behavioral pattern recognition is improved, but data storage requirements increase
Solution Approach 1:
The patent segments biometric data into discrete, standardized trajectory units that can be efficiently stored and processed. By dividing continuous behavioral data into discrete segments with standardized formats, the system enables long-term storage of behavioral patterns while optimizing storage efficiency through structured organization and compression.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There is provided a method and corresponding system for handling and/or generating anonymous biometric and/or behavioural data. The method comprises the steps of mapping (S1) biometric data originating from a subject into a biometric key using a one-way locality-sensitive hash function or receiving the biometric key. The method also comprises storing (S2) additional anonymous behavioural data bound to this key into an existing biometric trajectory in a database, wherein the behavioural data describes the user behavior of the subject. The method is performed to anonymize biometric data from a multitude of individuals, or subjects, per key, where each biometric key maps to biometric data of several subjects, and such a set of subjects resulting in the same biometric key is called a hash group, and a biometric trajectory is developed for each hash group.