Graph-based metadata modeling in indoor positioning systems

Abstract Modeling and persistence of different data structures in indoor positioning systems is a requirement for providing a large number of specialized location-based services. Collection and diversification of indoor positioning systems’ metadata are important to understand the context of the system’s operation to create a positive feedback improvement loop. While metadata from a residential building’s indoor positioning system operational context benefits the system (i.e. through occupancy patterns extraction that drive resource utilization strategies), it can also benefit the tenants’ well-being or drive other decisions through observing social dynamics. Observation of social relationships in residential buildings is rarely addressed due to highly stochastic movement patterns of tenants. In this article we have proposed a set of graph-based approaches for modeling social behavior data: modeling of tenants’ movement paths and detecting the existence of patterns, modeling of tenants’ social relationships (frequency, quality) as well as detecting social communities and tracking their evolution. We have tested our approaches on a real-world private residential building resulting in multidisciplinary implications and applications connecting the fields of IoT and indoor positioning to behavioral sciences. Finally, we provide public, high-quality positioning and occupancy datasets and open-source code for reproducing experiments on the observed residential building.