EVA STAR Trefferanzeige

Volltext thesis-submitted.pdf1.pdf (1,2 MB)
URN (für Zitat) http://nbn-resolving.org/urn:nbn:de:swb:90-308014
Titel Handling Live Sensor Data on the Semantic Web
Autor Hummel, Thomas
Institution Fakultät für Wirtschaftswissenschaften (WIWI)
Institut für Angewandte Informatik und Formale Beschreibungsverfahren (AIFB)
Dokumenttyp Buch
Verlag Karlsruhe
Jahr 2012
Hochschulschrift Abschlussarbeit - Bachelor
Abstract The increased linking of objects in the Internet of Things and the ubiquitous flood of data and information require new technologies in data processing and data storage in particular in the Internet and the Semantic Web.

Because of human limitations in data collection and analysis, more and more automatic methods are used. Above all, these sensors or similar data producers are very accurate, fast and versatile and can also provide continuous monitoring even places that are hard to reach by people.

The traditional information processing, however, has focused on the processing of documents or document-related information, but they have different requirements compared to sensor data. The main focus is static information of a certain scope in contrast to large quantities of live data that is only meaningful when combined with other data and background information.

The paper evaluates the current status quo in the processing of sensor and sensor-related data with the help of the promising approaches of the Semantic Web and Linked Data movement. This includes the use of the existing sensor standards such as the Sensor Web Enablement (SWE) as well as the utilization of various ontologies.

Based on a proposed abstract approach for the development of a semantic application, covering the process from data collection to presentation, important points, such as modeling, deploying and evaluating semantic sensor data, are discussed.

Besides the related work on current and future developments on known diffculties of RDF/OWL, such as the handling of time, space and physical units, a sample application demonstrates the key points.

In addition, techniques for the spread of information, such as polling, notifying or streaming are handled to provide examples of data stream management systems (DSMS) for processing real-time data.

Finally, the overview points out remaining weaknesses and therefore enables the improvement of existing solutions in order to easily develop semantic sensor applications in the future.