Performing data analysis on linked and open data in Smart Cities
Keywords: Linked Data, Event Streams, Shapes, Big Data Science, Internet of Things (IoT), Linked Open Data (LOD)
Supervision: Pieter Colpaert
Students: max 2
According to Juniper Research, the number of Internet of Things (IoT) devices will rise from 46 billion in 2021 to 83 billion by 2024. All those devices generate data and that data is sent over the internet to be stored in their respective cloud/data dump. Some of that sensor data is open data, which means that everybody can retrieve the data, make analyses and build applications on top of that data. An example of such open data is blue bikes in Ghent. It is possible to retrieve datasets with an open format, indicating that this information is available as 3-stars out of 5 on the 5 stars of open data. With only a few steps, it is possible to have those datasets as 5 stars. Those steps are using a mapping language (e.g. RML Mapper or Streamer) to transform the data to Linked Data. Furthermore, in order to publish it in a scalable, fragmented way, some extra information can be added by storing it as a Linked Data Event Stream (LDES). Then this collection is fragmented on a time based manner. In order to do an analysis on this collection of data, there is a need to extract a part of that collection. It is possible for each application to write its own method to extraction method, but it is more interesting to do it with a generic approach.