Composite User Profiles for Shared-Device Session Personalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing artificial intelligence systems struggle with generating personalized sessions on shared devices due to the complexity of obtaining high-quality data, the need for specialized knowledge to design and integrate AI solutions, and the difficulty in reviewing AI results, leading to challenges in determining user profiles accurately.
Innovation Solution
A system that selects between locally sourced and aggregated user profiles based on authentication types, using machine learning algorithms to assign weights to profiles, ensuring relevant content is presented to the user by distinguishing between inherence-based, knowledge-based, and possession-based authentications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a local user profile is used to present content on a shared device, then the content presentation is simplified, but the applicability of presented content to the user is reduced
Solution Approach 1:
The patent merges local user profiles with aggregated user profiles to create a composite user profile. The local profile contains user-specific data from the current device, while the aggregated profile contains data from multiple devices. By combining these profiles and weighting them based on authentication type, the system achieves both simplicity (using local profile) and accuracy (incorporating aggregated data), resolving the contradiction between ease of operation and content applicability.
2Reliability
If an aggregated user profile is used to present content on a device with one primary user, then the content applicability is improved, but the system complexity increases
Solution Approach 1:
The patent changes the parameter of profile composition based on authentication type. For biometric authentication (inherence-based), the system uses primarily local profile data with higher weight. For knowledge-based or possession-based authentication, the system incorporates aggregated profile data with higher weight. This dynamic parameter adjustment allows the system to maintain simplicity for single-user devices while providing accuracy for shared devices, resolving the contradiction between content applicability and system complexity.
3Reliability
If AI models are used to generate user profiles, then personalized content can be provided, but the difficulty in reviewing and improving models increases
Solution Approach 1:
The patent introduces an intermediary layer of weighted profile composition that mediates between raw AI model outputs and final personalized content. Instead of directly using complex AI models to generate all profile data, the system uses authentication type as an intermediary to select and weight appropriate profile sources (local vs. aggregated). This intermediary approach simplifies model review and improvement while maintaining profile accuracy, resolving the contradiction between profile accuracy and model review difficulty.
Data Source
AI summary
Methods and systems are described herein for generating composite user profiles using local profiles and aggregated user profiles. As one example, the system may present relevant data to users who initiate sessions on a user device. Specifically, the system may provide relevant data and content to users on user devices where they are the only user. Additionally, the system may present users who initiate a session on a shared device, content that is relevant to the specific user without commingling data or content from other users of the shared user device. Methods and systems are described herein for receiving, from a user, a user input to initiate a session on a user device and generating for display, in a user interface on the user device, a personalized device session based on the user profile.