FraPPE: A vocabulary to represent heterogeneous spatio-temporal data to support visual analytics

Abstract

Nowadays, we are witnessing a rapid increase of spatiotemporal data that permeates different aspects of our everyday life such as mobile geolocation services and geo-located weather sensors. This big amount of data needs innovative analytics techniques to ease correlation and comparison operations. Visual Analytics is often advocated as a doable solution thanks to its ability to enable users to directly obtain insights that support the understanding of the data. However, the grand challenge is to offer to visual analytics software an integrated view on top of multi-source, geo-located, time-varying data. The abstractions described in the FraPPE ontology address this challenge by exploiting classical image processing concepts (i.e. Pixel and Frame), a consolidated geographical data model (i.e. GeoSparql) and a time/event vocabulary (i.e. Time and Event ontologies). FraPPE was originally developed to represent telecommunication and social media data in an unified way and it is evaluated modeling the dataset made available by ACM DEBS 2015 Grand Challenge.