System And Method For Implicit Authentication

A method of authenticating a user on a mobile device, the method comprising:

• displaying first prompts on a display of the mobile device, the first prompts instructing a user in a first performance of a task, the task comprising a performance of an activity dependent on muscle memory;
• recording first sensor data obtained from a plurality of sensors incorporated within the mobile device, the sensor data obtained during the first performance;
• processing the first sensor data to determine first parameters of a predetermined model of the activity and storing the first parameters in a user model associated with the activity and the user, the user model being stored in a database within the mobile device;
• displaying second prompts on the display, the second prompts instructing the user in a second performance of the task;
• recording second sensor data obtained during the second performance;
• comparing the second sensor data to the user model to generate an authentication decision;
• and
• processing the second sensor data to determine second parameters of the predetermined model of the activity and storing the second parameters in the user model associated with the activity and the user, the second parameters updating the user model.

A system for implicit authentication for a mobile device associated with a user, wherein an authentication task is selected so as to leverage the user's muscle memory, wherein the mobile device comprises

• a touchscreen;
• a transaction authentication information unit;
• a sensor coupled to the transaction authentication information unit, the sensor used to obtain and transmit a set of data to the transaction authentication information unit, the set of data associated with a performances of the authentication task by the user, the set of data comprising data related to the tracking of a user's input on the touchscreen and the movement of the mobile device during the authentication task;
• and
• an anomaly detector coupled to the transaction authentication information unit, the anomaly detector generating an authentication model using the data set transmitted to the transaction authentication information unit, the generating comprising constructing a set of features for a user profile associated with the performance of the authentication task, and a parameter associated with the set of features, the anomaly detector performing training using the data set transmitted to the transaction authentication information unit, the training comprising:
• learning and adjusting the parameter associated with the set of features, the learning performed using a learning algorithm and storing the user profile associated with the adjusted parameter in a database;
• the anomaly detector authenticating the user using the data set transmitted to the transaction authentication information unit, wherein the anomaly detector analyzes the data set corresponding to the performance of the authentication task using an anomaly detection algorithm together with the stored user profile;
• and
• based on said analyzing, the anomaly detector authenticating the mobile device.