Querying Sensor Networks Using Temporal Property Graphs.

In this paper, sensor networks are considered, which are (stable) transportation networks equipped with sensors. We abstract transportation networks as graphs with sensor measurements that can be attributed to the nodes and edges. A sensor network is thus modelled as a graph with properties that can change throughout time, and which can be viewed as time series. To the best of our knowledge, a model, in which graph databases and time series analysis are combined to create a temporal property graph model, is new. This work also describes a language to query such temporal property graphs and discusses how both can be implemented and realised by using Neo4j and its query language Cypher. In short, this paper presents a database system for storing and querying sensor networks that enables future projects to reduce set-up times and prevent use-case specific implementation of database and querying applications.