An Ontology-based Approach to Represent Trajectory Characteristics
Résumé
The behavior of moving objects has been a relevant source of information to intelligent mobile systems. However, most of existing works on trajectory representation deal only with basic characteristics of trajectories, such as space and time, while these attributes may be not enough to provide the required information to intelligent systems. We observe that the analysis of other characteristics (e.g. speed and acceleration) of mobile objects enriches the trajectory description as well as open opportunities to novel applications. However, the dynamic nature of these characteristics brings several challenges related to the preprocessing and analysis of raw data. In this paper, we show how these additional characteristics may be integrated in trajectory modeling. We address the problem of representing trajectories with qualitative descriptions of movement modeled as an ontology. We validate our approach with real data from a sport tracking application.