Dependency Mining In Iot - From Research To Practice In Intelligent Transportation Systems

Large-scale Internet-of-Things (IoT) environments such as Intelligent Transportation Systems (ITS) are characterized by massive heterogeneities. Hence, cross-system monitoring of the underlying IoT-based Operational Technology (OT) is challenging, particularly due to the lack of information about functional dependencies between OT objects. We put forward a novel semantic-driven mining method aiming to identify functional dependencies out of IoT log files. In the current paper, we put this method into practice by (i) applying our approach to the domain of ITS, discussing it’s applicability step-by-step on the basis of a real-world example, (ii) putting forward a prototypical implementation giving details about the technologies used, and (iii) providing a systematic evaluation of our method from different perspectives in terms of accuracy, efficiency and effectivity grounded on simulated real-world data.