Privacy-Preserving Machine Learning with Secure MPC for Content Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning models trained on data from multiple sources face challenges in preserving user privacy due to the use of third-party cookies, which can lead to intrusive data access and loss of functionality when cookies are blocked.
Innovation Solution
A privacy-preserving machine learning platform utilizing secure multi-party computation (MPC) to categorize users into demographic groups without third-party cookies, enabling secure reporting of campaign effectiveness by maintaining user profiles at client devices and using encrypted data within a secure MPC cluster.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If third-party cookies are used to track user browsing behavior and categorize users into demographic groups, then digital component distribution and reporting effectiveness can be optimized, but user privacy is compromised and data security is weakened
Solution Approach 1:
The patent introduces a trusted intermediary system that acts as a mediator between users and digital component providers. This intermediary collects and processes user browsing data, categorizes users into demographic groups, and provides aggregated analytics to providers without revealing individual user identities. The intermediary uses techniques like data aggregation, anonymization, and secure multi-party computation to protect user privacy while enabling effective digital component distribution and reporting.
2Adaptability or versatility
If third-party cookies are used to trace user browsing history across websites, then relevant content can be displayed to users, but functionality is lost when cookies are blocked
Solution Approach 1:
The patent segments the tracking functionality into multiple components: (1) user profile data collected and stored locally in the user's browser, (2) demographic category identifiers generated and stored locally, and (3) aggregated analytics data transmitted to digital component providers. This segmentation allows the system to maintain content relevance through local user profiling while eliminating dependency on third-party cookies, ensuring functionality even when cookies are blocked.
3Measurement precision
If user data is collected and stored centrally to enable demographic-based digital component distribution, then reporting accuracy can be improved, but data security and memory usage are worsened
Solution Approach 1:
The patent extracts and stores only essential demographic category identifiers and aggregated analytics data in the user's browser, rather than storing complete user profiles or raw browsing data. The system extracts minimal necessary information (demographic categories) that enable accurate reporting and digital component distribution while significantly reducing data storage requirements and improving security by limiting the amount of sensitive data retained.
Data Source
AI summary
This disclosure relates to systems and techniques that can be implemented by content platforms to optimize (a) demographic-based digital component distribution used to categorize each user into a particular demographic so as to appropriately target that user for purposes of maximizing the efficacy of digital components shown to that user, and (b) demographic reporting used to report to digital component providers the effectiveness of the digital component.