Fuzzified semantic WEB reasoning for activity detection in WMSN applications

Thesis Type: Postgraduate

Institution Of The Thesis: Orta Doğu Teknik Üniversitesi, Faculty of Engineering, Department of Computer Engineering, Turkey

Approval Date: 2019


Consultant: ADNAN YAZICI


Activity detection in WMSNs is a hot topic for surveillance applications since automated determination of activities is a difficult process. This study aims to increase the reliability of activity detection by using semantic Web technologies extended with fuzzy logic. The proposed approach consists of three layers: sensor, data and semantic Web layers. The sensor layer includes a WMSN including sensor nodes with multimedia and scalar sensors. The data layer retrieves and stores data from the sink of the WMSN. At the top of the architecture, there is a semantic Web layer including a semantic Web application server, a fuzzy reasoning engine, and a semantic knowledge base. When there is a new entity detection at the sensor layer, the related data produced by the sensors and the sink is collected in the data layer and transmitted to the semantic Web application server where the data is converted to subjects-predicates-objects in accordance with designed ontology and saved in RDF format. Then, the fuzzy reasoning engine is automatically activated and fuzzy rules are executed to decide whether there is an activity in the controlled area. The proposed approach is implemented for an example surveillance application in which various threat types are deduced (i.e., attack, protest or ordinary mobility). In the implementation, the sensor layer is simulated with an application which produces synthetic data according to given scenarios. Users of the system can monitor occurring events in near-real time or replay recent events on their browsers. This implementation proves that the semantic Web technologies extended with fuzzy logic may have a significant impact on activity detection in WMSNs.