Machine Learning Query Generation with Evolving Knowledge Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data sharing techniques, such as Fully Homomorphic Encryption (FHE), are ineffective for secure data exchange among multiple organizations as they require key sharing, compromising data privacy and security.
Innovation Solution
A data exchange protocol that allows data sharing between organizations without decryption or key sharing, using symmetric keys and a dynamic knowledge graph to generate queries based on domain-specific events, ensuring data privacy through encoded responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Fully Homomorphic Encryption (FHE) is used for data encryption during computation, then data security is improved, but key sharing is required which compromises data privacy among multiple organizations
Solution Approach 1:
The patent introduces a trusted third party (TTP) as an intermediary that generates and distributes secret shares to multiple organizations. This TTP enables secure multi-party computation by allowing organizations to perform computations on encrypted data without needing to share their FHE keys directly, thus resolving the key sharing problem while maintaining data security
Solution Approach 2:
The patent segments the encryption key into multiple secret shares distributed among different organizations. Instead of requiring one organization to hold the complete FHE key, the key is divided into shares that can be used collectively for encrypted computation, eliminating the need for direct key sharing between organizations
2Reliability
If data is encrypted during transit and at rest, then data privacy is improved, but data needs to be decrypted for computation which creates security vulnerabilities
Solution Approach 1:
The patent merges the capabilities of encryption and computation by implementing a system where data remains encrypted throughout the computation process. By combining homomorphic encryption with secure multi-party computation, the system allows arithmetic operations to be performed directly on ciphertext without decryption, thus maintaining data privacy while enabling computation
3Productivity
If data sharing is increased among organizations, then business value is improved, but maintaining data privacy becomes more difficult
Solution Approach 1:
The patent implements dynamic access control where encryption keys and decryption permissions can be adjusted based on the specific computation task and participating organizations. This dynamic key management system allows flexible data sharing arrangements where organizations can access and compute on encrypted data from other organizations without permanently compromising data privacy, enabling both high data sharing and strong privacy protection
Data Source
AI summary
A method and system for generation of queries by machine learning (ML) models is provided. An ML model may generate a query based on reception of a data trigger. The query may be generated based on corresponding domain and a knowledge graph. The ML model may receive a response in an encoded format. The knowledge graph may evolve based on the response. A first subsequent query may be generated by the ML model based on the response and the evolved knowledge graph. The ML model may receive a response for the subsequent query in the encoded format. The ML model may determine whether the response culminates a current iteration. A second subsequent query may be generated by the ML model when the response does not culminate the current iteration. The current iteration may be terminated when it is determined that the response culminates the current iteration.