Discovering Process Models of Activities of Daily Living from Sensors

In recent years, more and more effort was put in the design and development of smart environments, which are aimed at improving the life quality of people, providing users with advanced services supporting them during their daily activities. In order to implement these services, smart environments are equipped with several sensors that continuously monitor the activities performed by a user. Sensor data are activation sequences and could be seen as the execution of a process representing daily user behaviors and performed activities. In this paper we propose a methodology, which exploit Process Mining techniques to discover both the daily behavior model and macro activities models. The former represents the “standard” behavior of the user in the form of a process model. The latter is a set of process models representing the flow of sensors activations when given tasks or macro activities are performed. A real-world case study is introduced to empirically show the efficacy of the proposed methodology