Entity Identity Generation via Deterministic Probabilistic Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for uniquely identifying entities across multiple devices and application domains face challenges, such as privacy concerns, instability in mobile device identification, and inability to track user interactions across different devices and platforms, particularly in non-browser-based applications.
Innovation Solution
A method and system that generate a unique entity identity by combining deterministic and probabilistic rules applied to feature data, including device and usage-specific information, to identify entities across multiple devices and platforms, allowing for customized content delivery and user data integration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If device-specific identifiers (UDID, Android_ID, MAC address, IMEI) are used to identify devices, then device identification uniqueness is improved, but privacy and security concerns worsen
Solution Approach 1:
The patent introduces an intermediary identifier system that mediates between device-specific identifiers and application needs. Instead of applications directly accessing device identifiers, an intermediary layer (application-generated identifiers like openUDID, SecureUDID, ODIN) is used to represent devices. This intermediary preserves device identification capability while reducing privacy concerns by not exposing actual device identifiers.
Solution Approach 2:
The patent extracts the identification function from device-specific identifiers and creates separate application-generated identifiers. By taking out the identification capability from the device layer and placing it at the application layer, the system maintains device identification uniqueness while eliminating the privacy and security issues associated with direct device identifier access.
2Object-affected harmful factors
If application-generated identifiers (openUDID, SecureUDID, ODIN) are used to overcome device identifier limitations, then privacy concerns are reduced, but accessibility to these identifiers is restricted to only applications with permission
Solution Approach 1:
The patent creates a universal identifier system that can be accessed by multiple applications through a standardized interface. The application-generated identifiers are designed to be universally accessible within the permission framework, allowing different applications to use the same identifier system without requiring special permissions for each application, thus achieving both privacy protection and ease of access.
3Measurement precision
If device-specific identifiers are used, then device-level identification is improved, but the ability to identify users across different devices deteriorates
Solution Approach 1:
The patent transitions from device-centric identification to user-centric identification by introducing a new dimension of identifiers that follow users across devices. Instead of identifiers being tied to specific devices (one dimension), the system creates identifiers that operate across the device dimension, enabling cross-device user identification while maintaining device-level precision through the multi-dimensional identifier framework.
4Productivity
If cookie-based techniques are used for user identification, then web-based user tracking is improved, but reliability in mobile applications and cross-device tracking deteriorates
Solution Approach 1:
The patent changes the fundamental parameters of identification from browser-based cookies to application-based identifiers. By changing the storage location (from browser to application), the identification mechanism becomes reliable across different browsers and devices. The identifier parameters are designed to persist across device changes while maintaining user association, solving the reliability issue in mobile applications and cross-device tracking.
Data Source
Figure 1
Figure 2~3
AI summary
An entity (a device, a user of a device or set of devices, a user of one or more applications on a device, a group of users of the device or set of devices, or the like) is identified across multiple device, usage, and application domains. The entity is assigned a unique entity identity that is generated from a set of feature data that model the entity. The feature data typically includes deterministic data, device and system-specific feature data, and usage feature data. The identity is generated by applying to the feature data one or more rules that identify which of the feature data to use to generate the entity identity. The rules include at least one deterministic rule, and at least one probabilistic rule. Periodically, an identity is merged into one or more entity identities that are found, by applying a rule, to represent a same entity.