Identifying Variation in Personal Daily Routine Through Process Mining: A Case Study

The study of daily routines has gained substantial attention, especially in healthcare. Understanding the activities and behaviors of individuals, particularly older adults, has the potential to play a crucial role in providing effective care and support, for example, when it comes to spotting deviations from it automatically. Process mining is a valuable tool for analyzing routine dynamics and identifying variations. However, human behavior is unstructured and characterized by variability, making it difficult to derive a process model representing only the control flow. In this paper, we employ a multi-dimensional process discovery and conformance checking methodology to a real-world dataset representing a person’s behavior in a smart environment. The derived model combines control flow and statistics on the data. The results, on the real-world data, highlight that the approach can identify variations in the inhabitant’s behavior.