Design And Implementation Of Two Semantic Enhancement Methods For A Sensor Application Platform

The project of this thesis involved is from the EU project ENVISION. Themain purpose is to provide two enhance approaches to the sensor observationplatform SensApp. In recent years, as the development of Geographical sensors, thesensors play an important role in our daily life. For example, the sensors can detectthe humility of the air conditions, and it can detect the flood possibility. So it isnecessary to provide such a platform. The SensApp is already developed but nowthe SensApp can only receive JSON data and SenML (Sensor Makeup Language)format, which is a sensor markup language that cannot provide semantic meaning.Thus it is impossible to make reasoning and use query language based on semantic,it cannot use various standards based on semantic to describe and retrieve sensors aswell. So it has real value to add the semantics to sensors to enhance SensApp.The sensor with semantic allows for both sensor descriptions and sensorobservation data with Semantic Web languages, that enables more expressiverepresentation, advanced access, and formal analysis of sensor resources. Thesemantic web annotates sensor data with spatial, temporal, and thematic semanticmetadata. This technique builds on current standardization efforts within the OpenGeospatial Consortium (OGC) Sensor Web Enablement (SWE), which includesSensor Observation Service (SOS), Sensor Model Language (SensorML),Observation&Measurement(O&M), and one semantic sensor network ontology(SSN ontology). They consist of the semantic web.The purpose of this thesis is to add semantic to sensors. Two enhanceapproaches can be given. The first one is to model one ontology and do thereasoning and query based on it, while the second approach is based on the reuse ofOGC SWE standards, to model the sensors according to the SensorML and O&Mand put the metadata and observation value to web services such as SOS, then makethe semantic annotation and finally discover the sensors information. By the use ofthese two approaches, the sensors can have semantics then enhanced descriptionsand access to the retrieve of sensor data can be provided.