The Benefits of Sensor-Measurement Aggregation in Discovering IoT Process Models: A Smart-House Case Study

IoT systems collect and exchange data whose analysis opens up incredible opportunities to improve the human satisfaction with IoT systems. The IoT data can be indeed used to discover human habits and interaction patterns, useful to both improve human experience and further automatize the system. Process Mining can be leveraged on for this purpose, but a gap needs to be bridged between IoT-device event data and logs by aggregating events to take to the right granularity for Process Mining. This papers reports on the experience on real-life data to discover the human habits in a smart house. In particular, the benefits are reported on how to aggregate event data to the right granularity to further apply process-mining discovery techniques. The results illustrate that, when applied on the case study, the proposed technique is able to discover human-habit models that are more readable and accurate, thus providing actionable insights for a subsequent optimization of the human experience with the IoT system.