Toward a Lightweight Ontology for Privacy Protection in IoT

The literature asserts that the design of an ontology-based privacy model is an essential starting point to address privacy risks in IoT, where connected devices are increasingly capable of monitoring human activities. Due to the omnipresence of data privacy concerns in IoT, we highlight the need for privacy ontologies that combine an expressive vocabulary with extension points but that do not overload the processing of privacy policies data. This paper presents IoT-Priv as a lightweight privacy layer upon IoT basic concepts such as device, sensor, and service. We introduce privacy requirements guiding the IoT-Priv ontology design, match these requirements to the respective privacy terms modeled, and show how to use IoT-Priv through a usage scenario. Finally, we evaluate static metrics and response times of spatial and temporal query filters over instances of privacy policies. Results open the way for the creation of scalable, privacy-enabled systems.