Inhabitant identification from activity patterns using a graph-based data mining technique

The goal of the MavHome smart home project is to build an intelligent home environment that is aware of its inhabitants and their activities. Such a home is designed to provide maximum comfort to inhabitants at minimum cost. This can be done by learning the activities of the inhabitants and automating those activities. Different inhabitants in the home can follow different activities. Therefore to achieve this goal, it is necessary to identify among multiple inhabitants who is currently present in the home. MavHome data is structural in nature and hence we use a structural approach for inhabitant identification. Subdue is a graph-based data mining algorithm that discovers patterns in structural data. By representing the activity patterns for each inhabitant as a graph, Subdue can be used for inhabitant identification. We introduce a multiple-class learning version of Subdue and validate this approach through experimentation on synthetic as well as real smart home activity data for multiple inhabitants.