Automated Login Method Selection in a Multi-modal Authentication System : Login Method Selection based on User Behavior

摘要

An intelligent model is proposed to automate selection of login authentication method in a multi-model authentication system based on user behavior profile. The proposed protocol analyzes the user behavior data to minimize the process in real time and prevent untrusted attempt, meanwhile, it able to facilitate frictionless user experience in a multi-modal authentication system. The proposed system determines a user at a log-in interface and retrieves the user's behavioral historical data as its determinant factors, matching user behavior based on data retrieval from database. Wherein the behavioral data includes the user's login sessions time, geolocation, accessed application and type of user-agent used. The system processes the behavioral data to generate a user profile by a profiling module, at that point, evaluates the user's profile by factoring in environmental parameters of the user terminal to select an authentication method by an evaluation module. The selected authentication method will be displayed for the user to complete the authentication process. On the other hand, the system will updating the user profile with data relating to a new successful login session for future evaluation.