Data De-identification via Event Fragment Sequence Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing k-anonymity technology is inadequate for processing continuous data, such as time-series data, and may result in data loss and risk of re-identification, especially in scenarios like Internet trading and Internet of Things (IoT), where protecting privacy is crucial.
Innovation Solution
A data de-identification method that involves obtaining event fragment sequences from original data based on an event condition, adjusting sequence data to ensure that at least k-1 items of data cannot be differentiated, using techniques like Apriori or PrefixSpan algorithms to manage sub-sequences and frequencies, thereby enhancing privacy protection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing k-anonymity technology is used to de-identify data, then personal identity protection is improved, but data completeness deteriorates and re-identification risk increases
Solution Approach 1:
The patent segments continuous data into discrete event fragments with specific characteristics (event type, time, location, etc.). By dividing the data into structured event fragments and analyzing their sequence patterns, the method maintains data completeness while enabling effective de-identification through sequence frequency analysis rather than direct k-anonymity application on continuous data.
2Reliability
If existing k-anonymity technology is used to de-identify data, then personal identity protection is improved, but the risk of re-identification increases
Solution Approach 1:
The patent applies dynamic sequence frequency adjustment by analyzing the frequency of event fragment sequences and dynamically modifying sequences that fall below the k-anonymity threshold. Low-frequency sequences are merged or modified to match higher-frequency patterns, creating a dynamic adaptation process that continuously maintains k-anonymity while preserving data utility, thereby reducing re-identification risk.
3Reliability
If event fragment sequences are extracted and adjusted to meet k-anonymity requirements, then re-identification risk is reduced, but data processing complexity increases
Solution Approach 1:
The patent performs preliminary extraction of event fragments and their sequences before applying de-identification adjustments. By pre-identifying all event fragments, their characteristics, and sequence patterns in advance, the system can efficiently determine which sequences need modification and apply adjustments systematically, reducing overall processing complexity compared to iterative approaches.
Data Source
AI summary
A data de-identification method, a data de-identification apparatus and a non-transitory computer readable storage medium executing the same are provided. Original data including an identification field, a condition field, and a record field is obtained. An event condition is obtained according to the condition field. From the original data, a plurality of event fragment sequences corresponding to each of a plurality of identification data and corresponding to the event condition are obtained according to the plurality of identification data in the identification field and the event condition. Sequence data is obtained according to the plurality of identification data and the plurality of event fragment sequences corresponding to each of the identification data. De-identification data is obtained by adjusting the sequence data.