Mining Contextual Process Models Using Sensors Data: A Case of Daily Activities in Smart Home

Different techniques are used by companies to enhance their processes. Process mining (PM) is one of these techniques that relies on the user activity logs recorded by information systems to discover the process model, to check conformance with the prescribed process, to enhance the process, and to recommend or guess the next user activity. From another hand, many contextual factors such as time, location, weather, and user's profile influence the user activities. However, PM techniques are mainly activity-oriented and do not take into consideration the contextual environment. Our main goal is to enrich process models obtained using process mining technics with contextual information issued from sensors data and to construct contextual process models for a better process discovery, conformance checking, and recommendations. In this paper, we test the feasibility to integrate events logs with sensor logs to provide meaningful results. We use existing datasets with events and sensors logs about daily activities in Smart Home to construct a process model enriched by contextual information.