Semantic modeling of the platform
Over the last years, advancement in wireless sensor networks has increased the interest in paradigms to integrate data and capabilities provided by physical world objects on virtual representations. According to this, Semantic technologies can assist in managing, querying, and combining sensors and observation data. Semantic technologies allow users to work with abstraction levels above the technical details of format and integration.
In accordance with the requirements and cases of use defined within the scope of the Citisim project, an ontology-based on the SSN ontology[1] has been proposed. This ontology uses only some of the modules provided by SSN. However, if new requirements are originated, the rest of the modules could be used to extend the proposed ontology. The proposed ontology is presented below.
As can be seen in the figure, the proposed ontology is formed of eleven classes. These classes belong to four ontologies. First, the modeling of sensors (sosa:Sensor) and observations (sosa:Observation); the actuators (sosa:Actuation) and actions (sosa:Actuation); and the properties with which sensors and actuators interact (sosa:ObservableProperty, sosa:ActuableProperty, sosa:FeatureOfInterest) are defined in the SOSA ontology.
Moreover, the platforms (sosa:Platform) where sensors and actuators are hosted are modeled in SOSA ontology. Second, the modeling of systems (ssn:System) is defined in SSN ontology. Finally, the results of observations and performances given as integers and their time stamp are modeled with RDF and XSD ontologies. In addition to these classes, the relationships between instances shown in the image are also considered.
This ontology has been used for modeling the sensors integrated into the Citisim projects, as well as their data. With this purpose, a system for modeling this information has been developed. This system gets information about the sensors (including the different events with the measures sent by them) and models it according to the proposed ontology. Then, this information is stored in a semantic repository, allowing the retrieval and filtering of the information using semantic technologies. This enables to take advantage of the linked data benefits to query data from other semantic repositories.
[1] https://www.w3.org/TR/vocab-ssn/