Global Behavioral Verification Model for Mobile Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current user verification systems for mobile devices face memory overload and inefficiencies due to the need for individual models for each user, which complicates debugging, updating, and analysis, and fail to continuously authenticate users throughout sessions, leaving them vulnerable to attacks.
Innovation Solution
A method and system that uses behavioral embedding to generate a global verification model, allowing for user verification without building separate models for each user, by extracting features from interaction data, aggregating them into feature vectors, and projecting them into a global feature space for continuous authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual verification models are built for each user, then user-specific verification accuracy is improved, but memory consumption and system complexity increase
Solution Approach 1:
The patent merges individual user verification models into a single global verification model that processes all user interactions. Instead of maintaining separate models for each user, the system aggregates interaction data from multiple users and trains one unified model, thereby reducing memory consumption while maintaining verification capabilities through shared behavioral patterns
Solution Approach 2:
The global verification model is designed to be universal across all users, performing the same verification function for any user without requiring user-specific model instances. This multi-functional approach allows the single model to adapt to different users' behavioral patterns while consuming constant memory resources
2Adaptability or versatility
If individual verification models are built for each user, then personalized verification is improved, but model maintenance and debugging difficulty increase
Solution Approach 1:
By combining all user verification logic into a single global model, the system eliminates the need to maintain, update, and debug multiple separate models. Model maintenance tasks such as retraining, hyperparameter tuning, and bug fixes need to be performed only once on the global model rather than replicated across numerous user-specific models
Solution Approach 2:
The global verification model achieves personalized verification by dynamically adjusting its internal parameters based on user-specific interaction patterns. Instead of creating different models for different users, the system modifies the model's learned parameters through training on user-specific data, maintaining personalization capability while simplifying model management
3Reliability
If traditional authentication mechanisms are used, then initial authentication is achieved, but continuous authentication throughout sessions is lost
Solution Approach 1:
The verification model operates continuously throughout the user session, constantly analyzing interaction data to verify user identity. Instead of performing authentication only at login, the system maintains ongoing verification by processing each user interaction through the model, ensuring continuous security monitoring for the duration of the session
Solution Approach 2:
The system implements continuous feedback by constantly comparing actual user interactions against the verification model's expectations. Each interaction provides feedback that reinforces or challenges the current authentication state, allowing the system to detect anomalies and potential credential theft in real-time throughout the session
Data Source
AI summary
User verification using a mobile interaction can include receiving interaction data associated with a user of an application operating on a mobile device, the interaction data is generated during an interactive session of the user with the application on the mobile device. A plurality of features are extracted from the interaction data and one or more feature vectors are generated from the plurality of features. The plurality of features are aggregated to the one or more feature vectors and embedded within each feature vector. The embedded plurality of features are then projected to a global feature space by comparing a history of interactive sessions associated with the user and a history of interactive sessions associated with a plurality of different users for generating a global training dataset. Finally, a verification model is generated as a global binary classification model using the global training dataset.