Probabilistic Framework for Device Association
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud platforms face challenges in determining associations between devices used by a single user, as different devices employ unique identifiers and features, making it difficult to track user information and habits across multiple devices.
Innovation Solution
A probabilistic framework using a hinge-loss Markov Random Field (HL-MRF) model and probabilistic soft logic (PSL) is employed to analyze device characteristics and connection information, estimating probability density functions to determine device associations based on machine-learning processes, and transmitting this information to associated devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If different identifiers and features are used for each device, then device uniqueness and identification accuracy are improved, but the ability to associate devices with a single user deteriorates
Solution Approach 1:
The patent combines multiple device-level identifiers and features into a unified user-level association model. By merging data from multiple devices through probabilistic soft logic and Markov random fields, the system recovers lost user association information while preserving device uniqueness.
Solution Approach 2:
The patent introduces probabilistic soft logic and Markov random field models as intermediaries between device identifiers and user associations. These mathematical frameworks serve as mediators that infer user-level connections from device-level data without requiring direct user identification on each device.
2Measurement precision
If device-level analysis is performed, then device-specific data accuracy is improved, but user-level analysis capability deteriorates
Solution Approach 1:
The patent transitions from device-level to user-level analysis by adding a temporal and probabilistic dimension. Through sequential Bayesian inference and Markov random fields, the system elevates device-specific data into user-level insights, enabling multi-device user behavior analysis while preserving device-level accuracy.
3Measurement precision
If multiple devices are monitored independently, then device monitoring precision is improved, but user behavior tracking across devices deteriorates
Solution Approach 1:
The patent establishes continuous user behavior tracking across devices through probabilistic models. By maintaining persistent user associations through time and across device boundaries using Markov random fields, the system ensures continuous observation of user behavior patterns without breaking the tracking chain when users switch devices.
Data Source
AI summary
Methods, systems, and devices for determining device associations are described. Some database systems may store information related to device characteristics. Each of these devices may be operated by one or more users, and each user may operate one or more devices. In some cases, information about users may be more valuable than information about devices. As such, a system may determine probable associations between devices, where an association can correspond to operation by a same user. To determine device associations, the system may perform a machine-learning process (e.g., using probabilistic soft logic (PSL) and a hinge-loss Markov Random Field (HL-MRF) model) on input device characteristics and connection information to generate a probability density function. The probability density function may indicate associations between devices within the system. Based on one or more thresholds, the system may determine sets of associated devices and may transmit this association information for analysis or display.