Encrypted Decision Tree Evaluation for Private Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Creating a classifier for continuous authentication requires large amounts of sensitive user data, leading to privacy and security violations if disclosed.
Innovation Solution
A three-party protocol using secure multi-party computation (MPC) with additive homomorphic encryption and oblivious transfer to encrypt user data and partial decision tree models, allowing classification without revealing sensitive information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large amounts of user data are used to create a classifier for continuous authentication, then authentication accuracy is improved, but privacy and security are violated due to data disclosure
Solution Approach 1:
The decision tree model is divided into two separate shares, with each server holding only a partial share. Neither server has access to the complete model, preventing full data disclosure while enabling collaborative authentication decisions through secure computation on encrypted data.
Solution Approach 2:
Homomorphic encryption serves as an intermediary mechanism that allows computation on encrypted data without decryption. The servers can perform authentication computations on encrypted user data and model shares, with only the final classification result being decrypted and shared with the client, thus preventing data disclosure throughout the computation process.
2Loss of information
If user data and decision tree models are encrypted using homomorphic encryption, then privacy is protected, but computational complexity increases
Solution Approach 1:
The computational workload is segmented and distributed across two servers, each performing computations on their respective model shares and encrypted data. This distribution reduces the computational burden on any single server and enables parallel processing of authentication decisions.
Solution Approach 2:
The patent combines homomorphic encryption with secure multi-party computation protocols to achieve both privacy protection and efficient computation. The encryption and computation operations are integrated into a unified protocol that leverages cryptographic properties to simplify the overall computational complexity.
3Reliability
If a three-party protocol with secure multi-party computation is implemented, then privacy-preserving authentication is achieved, but system complexity increases
Solution Approach 1:
The system is segmented into three independent parties: the client holding user data and two servers holding partial model shares. Each party operates independently with limited access to sensitive information, reducing the complexity of any single component while maintaining overall system security through distributed collaboration.
Data Source
AI summary
Disclosed herein are system, method, and computer program product embodiments for evaluating a decision tree in a privacy preserving manner. In embodiments, s first server receives a first partial share of the decision tree and encrypted input data from a client, wherein the input data comprises a set of attributes. A second server receives a second partial share of the decision tree. Each of the first server and the second server communicate with the other server to compute a classification result over the encrypted input data using the respective partial share of the decision tree received thereby and secure multi-party computation methods, including additive homomorphic encryption and oblivious transfer, such that the classification result is computed without decrypting the encrypted input data, and such that the classification result is ultimately received by the first server. The first server then transmits the classification result to the client.