Data Cleansing Pipeline for Secure Cross-Organization Computation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing secure computation systems struggle to perform analysis when data from different organizations is in varying formats, leading to incomplete or inaccurate results due to format discrepancies.
Innovation Solution
A processing system that includes a cleansing server to perform data cleansing processing, converting and aligning data formats, and a secure computation system to store and analyze data in a fragmented share state across multiple servers, ensuring secure and accurate cross-sectional analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data from multiple organizations with different formats is directly input into a secure computation system, then data security and personal information protection are maintained, but analysis cannot be performed due to format discrepancies
Solution Approach 1:
A data conversion unit is introduced as an intermediary component between data input and the secure computation system. This unit automatically converts data from various formats into a unified format required by the secure computation system, enabling format compatibility without compromising data security. The intermediary handles format standardization while the core secure computation system remains unchanged and secure.
2Adaptability or versatility
If data cleansing processing is performed on data from multiple organizations, then data format compatibility and analysis capability are improved, but system complexity increases
Solution Approach 1:
The data conversion unit is integrated into the existing secure computation system architecture, merging the data cleansing functionality with the core system rather than operating as a completely separate preprocessing system. This integration reduces overall system complexity by consolidating components and streamlining the data flow from input through conversion to secure computation.
3Productivity
If a unified data format is enforced across all organizations, then analysis processing can be performed smoothly, but ease of data contribution and organizational autonomy are reduced
Solution Approach 1:
The automatic format conversion is performed as a preliminary action immediately when data is submitted, before the data enters the secure computation system. The conversion unit automatically detects and transforms data into the required unified format without requiring manual intervention from data contributors, maintaining ease of data submission while ensuring format compatibility for efficient analysis processing.
Data Source
AI summary
A processing system includes a cleansing server that includes a cleansing unit that performs data cleansing processing for secure computation registration on pieces of data transmitted from a plurality of organizations, a secure computation system that stores the pieces of data after the data cleansing processing in a fragmented share state in a distributed manner across a plurality of servers, and performs analysis processing in response to an analysis request on secure computation based on the pieces of data after the data cleansing processing, and a web server that outputs an analysis result of the analysis processing performed in response to the analysis request to a request source of the analysis request.