A Knowledge Driven Policy Framework for Internet of Things

Enabling smart environments through scalable policy reasoning and Internet of Things

“…In order to address the aforementioned issues, we proposed a semantically‐aware policy framework (Goynugur et al, ) that can (a) define high‐level policies and refine them to device‐ and service‐ level policies in context; and (b) automatically detect policy conflicts and propose resolutions using an artificial intelligence (AI) planner; and (c) scale well in dynamic environments in terms of policy refinement, conflict detection, and resolution with minimal human intervention. The policy language is based on OWL‐QL (Motik et al, ) and it defines the concepts and relations of the target application domain.…”

“…We observed that many of today's IoT applications such as smart home (e.g., Amazon Echo, Google Home), intelligent manufacturing, or health and safety solutions operate by collecting data from sensors and devices into a central hub that makes the decisions. Motivated by this observation, we developed an ontology‐based policy framework to reason over policies so as to manage the behaviors of an IoT system in‐context using a well‐defined knowledge base (Goynugur, Mel, Sensoy, Talamadupula, & Calo, ). The use of ontologies enables us to maintain interoperability between heterogeneous devices and to define a high level vocabulary to author policies, while providing efficient mechanisms to do reasoning (i.e., inferring class memberships, detecting policy conflicts).…”

“…Unfortunately, OWL‐QL lacks expressivity to describe the most of the real life scenarios. The policy language of the above framework (Goynugur et al, ) does not allow comparing data properties in policy definitions. For example, it is not possible to represent a statement like “oxygenLevel ≤ 19%.” Furthermore, typical in‐use IoT solutions come with multiple heterogeneous databases (and/or software components) that rely on varying schemata and they implement their own communication protocols.…”