Entropy-Based Privacy Quantification for Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing privacy preservation techniques for sensor data often over-provision privacy, leading to irreversible utility loss of the data, as they consider worst-case scenarios without measuring the required privacy level, resulting in data distortion and loss of intelligence.
Innovation Solution
A method and system for privacy measurement and quantification of time-series sensor data using a privacy measurement factor calculated through entropy computation and statistical compensation, followed by scaling to determine a privacy quantification factor, which guides the application of appropriate privacy preservation techniques to ensure optimal data protection without unnecessary distortion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If arbitrary privacy preservation techniques are applied to sensor data, then privacy protection is improved, but data utility and intelligence are lost
Solution Approach 1:
The patent applies parameter changes by measuring the privacy level of sensor data using entropy computation and statistical compensation, then adjusting the privacy preservation strength according to the measured privacy level. This allows dynamic adjustment of privacy protection parameters rather than applying fixed arbitrary privacy preservation, thereby maintaining data utility while providing appropriate privacy protection.
2Reliability
If encryption is applied to sensor data, then privacy protection is improved, but complete utility of the data is destroyed
Solution Approach 1:
The patent applies partial action by measuring the actual privacy level required and applying only the necessary amount of privacy preservation technique. Instead of applying full encryption that destroys all utility, the system applies privacy preservation proportionally to the measured privacy level, preserving data utility while providing adequate privacy protection.
3Reliability
If strong privacy preservation is applied to sensor data, then privacy protection is improved, but sensor data becomes useless
Solution Approach 1:
The patent applies dynamics by making the privacy preservation strength adaptive based on the measured privacy level of the sensor data. The system dynamically adjusts the privacy preservation technique according to the actual privacy requirements, rather than applying static strong privacy preservation that would make data useless. This maintains data intelligence and adaptability while providing appropriate privacy protection.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
System(s) and method(s) to provide privacy measurement and privacy quantification of sensor data are disclosed. The sensor data is received from a sensor. The private content associated with the sensor data is used to calculate a privacy measuring factor by using entropy based information theoretic model. A compensation value with respect to distribution dissimilarity is determined. The compensation value compensates a statistical deviation in the privacy measuring factor. The compensation value and the privacy measuring factor are used to determine a privacy quantification factor. The privacy quantification factor is scaled with respect to a predefined finite scale to obtain at least one scaled privacy quantification factor to provide quantification of privacy of the sensor data.