Secured Virtual Containers for Private Hyper-Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users are unable to receive hyper-personalized experiences while retaining control over their private data, and service providers are hesitant to share proprietary machine learning models due to concerns about data misuse and security.
Innovation Solution
A secured virtual container is implemented on a computing device, isolated from the operating system, to store and process user data using a machine learning model, generating obfuscated inference values that are accessible only through a broker, ensuring data and model secrecy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If users provide all their private data to a vendor/service provider for hyper-personalized experiences, then the personalization accuracy is improved, but the data security and user trust deteriorate
Solution Approach 1:
The system divides the data processing function into two separate segments: a data collection component that gathers user data and a separate inference generation component that processes the data through machine learning models. This segmentation allows the system to achieve accurate personalization while maintaining security through functional separation, as the inference component operates independently from the data collection component.
Solution Approach 2:
The patent introduces an intermediary mechanism where inference values are generated and made available through a controlled interface rather than direct data access. The system generates inference values from user data and makes these values available to applications through a broker or controlled interface, preventing direct access to raw user data while still enabling personalized experiences.
2Adaptability or versatility
If service providers distribute proprietary machine learning models to achieve hyper-personalization, then the personalization capability is improved, but the model secrecy and proprietary intelligence deteriorate
Solution Approach 1:
The system extracts the machine learning model from the data processing pipeline and places it within a secured virtual container that is isolated from the operating system. This extraction allows the model to be used for personalization purposes while remaining protected within the container boundary, preventing unauthorized access or distribution of the proprietary model.
Solution Approach 2:
Instead of distributing the actual proprietary machine learning model, the system creates a copy or instance of the model within the secured virtual container. This copied model can perform inference operations locally without requiring distribution of the original proprietary model, thus maintaining model secrecy while enabling personalization functionality.
3Reliability
If a secured virtual container is implemented to protect user data and models, then the data privacy and model security are improved, but the system complexity deteriorates
Solution Approach 1:
The system implements a nested structure where the secured virtual container is contained within the operating system environment. The virtual container itself contains both the user data and the machine learning model, creating a nested arrangement where data and models are protected within a single isolated environment. This nesting approach provides comprehensive protection while organizing complexity in a hierarchical manner.
Data Source
AI summary
A secured virtual container is enabled to securely store personal data corresponding to a user, where such data is inaccessible to processes running outside the secured virtual container. The secured virtual container may also include an execution environment for a machine learning model where the model is securely stored and inaccessible. Personal data may be feature engineered and provided to the machine learning model for training purposes and/or to generate inference values corresponding to the user data. Inference values may thereafter be relayed by a broker application from the secured virtual container to applications external to the container. Applications may perform hyper-personalization operations based at least in part on received inference values. The broker application may enable external applications to subscribe to notifications regarding availability of inference values. The broker may also provide inference values in response to a query.