Smart Home System Using Sequential Pattern Mining

Abstract

Smart home systems are one of today's activeworking issues because of the increase comfort, raised living standards,security, health, economic benefits. Smart home systems should interact withthe activities of people living in a house and be remotely accessible,monitorable and controllable. A smart home system with the mentioned featureshas been developed in this study. The system collects data from the behaviorsand interactions of the people in the house. It evaluates the collected dataadaptively and outputs the results. It produces control signals by using theresults. In addition, a simulation was developed to collect data from devicessuch as an IoT device. The resulting data is examined using rule subtractionfrom sequential patterns, a data mining method. In this way, the habits of theusers in a house are added to the system as a scenario. By using scenarios andrules, the interaction with the user of the system is increased by predictingwhat the user can do next. The results show energy saving and comfort increasewith the developed system.