Federated Data Exploration System for Secure Cross-Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data sharing methods, such as file downloading and APIs, lead to loss of data ownership, difficulty in cross-analysis across multiple datasets, and inadequate access control, making it challenging for users to derive valuable insights from shared data.
Innovation Solution
A federated data exploration system with interoperable data exploration apparatuses that separate data ownership from usage, using a data process engine and visualization user interface to enable secure, granular access and cross-analysis by exchanging identifiers rather than entire datasets, ensuring data ownership protection and facilitating user-friendly data exploration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If data is shared via file downloading or APIs, then data accessibility is improved, but data ownership is lost and security control deteriorates
Solution Approach 1:
The patent segments data access into multiple levels: metadata access, identifier exchange, and selective data retrieval. Instead of sharing entire datasets, users exchange identifiers and retrieve only specific portions needed, maintaining ownership control while enabling accessibility.
Solution Approach 2:
The patent introduces an intermediary data exploration apparatus that mediates between data owners and data users. This intermediary enables secure data sharing by acting as a trusted third party that facilitates identifier exchange and controlled data retrieval without compromising ownership.
2Adaptability or versatility
If entire datasets are shared for cross-analysis, then analytical capability is improved, but data transfer time and storage requirements increase
Solution Approach 1:
The patent extracts only the essential elements (identifiers and metadata) needed for cross-analysis from complete datasets. Users can perform analytical operations by exchanging identifiers and retrieving only the specific data portions required, eliminating the need to transfer and store entire datasets.
Solution Approach 2:
The patent implements partial data retrieval where users obtain only the specific portions of datasets needed for their analysis rather than entire datasets. This partial action approach reduces data transfer time and storage requirements while maintaining analytical capability.
3Reliability
If fine granularity access control is implemented, then data security is improved, but system complexity increases
Solution Approach 1:
The patent segments access control into granular levels operating on different data objects, allowing independent control policies for each segment. This segmentation enables fine-grained security control while managing complexity through modular, hierarchical access management.
4Reliability
If data is shared from owner's viewpoint, then data protection is improved, but usability for users deteriorates
Solution Approach 1:
The patent introduces a data exploration apparatus as an intermediary that translates between owner's protective requirements and user's usability needs. This intermediary enables users to explore and analyze data in user-friendly ways while the system enforces ownership protection and access control policies.
Data Source
AI summary
A federated data exploration system includes a first and a second interoperable data exploration apparatuses. The first interoperable data exploration apparatus executes a data process engine and has a first data object. The data process engine enables a display screen to display a visualization user interface, which includes a first and second display areas respectively for the first and second data object. The second data object belongs to the second interoperable data exploration apparatus. The data process engine explores the second data object by issuing a sequence of queries in response to a sequence of operations on the second display area to narrow down a target of interest. The data process engine receives a first set of identifiers corresponding to the target of interest from the second data object and produces a target dataset by querying the first data object according to the first set of identifiers.