Direkt zu

Informationen für Studierende

zur Startseite


A Common Vocabulary for the Internet of Things
Betreuer Dr. rer. nat. Pascal Hirmer
Prüfer Prof. Dr.-Ing. habil. Bernhard Mitschang


In the Internet of Things, devices equipped with sensor and actuators interact with each other through standard internet protocols to reach common goals. For example, in the domain of Smart Homes, heating systems can be automatically regulated to increase their efficiency. Furthermore, there are many other domains that can be enriched with IoT technologies in order to enable automation. Common examples are Smart Factories, Smart Cities, Smart Offices, or Smart Energy Grids. The main approach to build IoT applications for these domains is extracting data from sensors of the environment, analyze their data and react on occurring situations, e.g., by invoking actuators. To extract sensor data and provision them to IoT applications, a large variety of IoT platforms were developed in the past using different protocols (MQTT, HTTP). Famous examples for such platforms are Mosquitto, FIWARE, or openmtc. A way to describe how the data originating from such platforms is structured, is the Topic Description language, which provides a means to describe the characteristics of the sensors and actuators, the structure of their data, and metadata, such as location or owner.
However, in order to describe data originating from IoT platforms, a common vocabulary is essential. For example, the same type of temperature sensor could be described as “Temp”, “T1”, “T12345”. Searching for specific sensors, data structures, locations, etc. can be very difficult if there is not common way to define the terms than can be used so that synonyms can be recognized.

In this Master thesis, the goal is developing a common vocabulary for the Internet of Things. This vocabulary has to be able to define common terms for specific IoT domains, e.g., names of sensors, and, furthermore, should be able to recognize synonyms between these terms. The main idea is using a semantic approach, e.g., using ontologies or RDF. Furthermore, familiarization with taxonomies is of vital importance of this thesis. The created solution needs to provide a suitable data store for the vocabulary, a common standardized API to access it, a Web UI to browse it (using a graph view) and make changes, and should be integrated with the TDLIoT. The concrete tasks are:

  • Thorough familiarization with related work (similar approaches, not only foundations)
  • Generic Concept for the vocabulary (semantics or not, structure, taxonomies)
  • Search for suitable technologies to realize the vocabulary backend (e.g., Apache Jena)
  • Design of the Web UI (browsing, graph-view, vocabulary, user management, …)
  • Search for suitable technologies to realize the vocabulary frontend (e.g., Angular, React, …)
  • Implementation of frontend and backend application
  • Integration with the TDLIoT