Group Profiles for Shared Device User Attribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing recommendation systems face challenges in accurately attributing customer data to the correct user account, especially in scenarios where multiple users share a common system or device, leading to difficulties in providing personalized item recommendations.
Innovation Solution
Implementing group profiles that track both individual user account data and group behavioral data to generate tailored item recommendations relevant to a group of users, using techniques such as device-enabled identification and automatic user detection to manage user accounts and provide content access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional user account systems are used where each user logs in individually, then accurate attribution of customer data to the correct user account is achieved, but the system becomes difficult to operate when multiple users share a common device and the user experience deteriorates due to constant login/logout requirements
Solution Approach 1:
The system automatically detects the presence of users through device sensors and cameras, and autonomously determines which user profile to activate without requiring manual login actions. The fulfillment system performs self-service user identification by analyzing biometric data and device information to attribute behavior data to the correct user account.
Solution Approach 2:
The patent replaces the mechanical login/logout process with automated optical and sensor-based detection systems. Instead of manual authentication mechanisms, the system uses camera-based facial recognition, device sensor data, and network information to automatically identify users and switch between profiles seamlessly.
2Ease of operation
If multiple user profiles are supported on a single device, then ease of operation is improved for shared devices, but accurate attribution of behavior data to the correct user account becomes difficult
Solution Approach 1:
The patent introduces device information as an intermediary element that links user profiles to the physical device. Each user profile is associated with specific device characteristics (camera data, sensor readings, network identifiers), allowing the system to accurately attribute behavior data to the correct user by matching current device state with stored profile information.
Solution Approach 2:
The system continuously monitors device sensors, camera input, and network state to provide real-time feedback about which user is currently present. This feedback loop allows the fulfillment system to dynamically adjust user profile activation and maintain accurate attribution of behavior data by comparing current detection data with stored user profile characteristics.
3Measurement precision
If manual user identification is required for each transaction, then attribution accuracy is maintained, but productivity decreases due to time-consuming authentication processes
Solution Approach 1:
The system performs preliminary user identification by continuously monitoring device sensors and camera data in the background, even before a purchase transaction occurs. User profiles are pre-loaded and ready for activation based on detected presence, eliminating the need for time-consuming authentication processes during actual transactions while maintaining accurate attribution.
Solution Approach 2:
The fulfillment system automatically performs user identification and profile switching without requiring manual authentication actions from users. The system self-services the attribution process by analyzing device data, camera input, and network information to determine which user account should be activated for the current transaction.
Data Source
AI summary
A network-based enterprise or other system that makes items available for selection to users may implement group profiles for group item recommendations. A request for item recommendations offered via the network-based enterprise may be received. Multiple user accounts of the network-based enterprise may be detected as associated with the request. A group profile that includes the detected user accounts may be identified. Group profiles may maintain behavioral data for the respective user accounts included in the group profiles, as well as behavioral data for the group profile as a whole. Item recommendations may be generated according to the group profile and provided to the detected users. Group profiles may available across multiple systems and devices so that item recommendations based on a group profile may be provided to different systems.