Context Data Preprocessing for Context-Aware Smartphone Authentication

This paper proposes an approach to carrying out context data preprocessing gathered from smartphone users to support context-aware authentication. Context-aware authentication is a technique to implicitly authenticate a smartphone user using contextual data (e.g., call log, location) without explicitly requesting the user’s any actions. In order to enable context-aware authentication, a user’s contextual data should be carefully processed for learning user’s past contextual patterns in consideration of user’s hourly, daily, weekly or monthly behaviors. In this paper, we gathered contextual data from 200 voluntary smartphone users for about 2 years and showed what the appropriate contextual data is preprocessing for performing context-aware authentication.