Abstract
Summarizing spatiotemporal trajectories of a large number of individual objects or events provides insight into collective patterns of phenomena. A well-defined data model can serve as a vehicle for classifying and analyzing data sets efficiently. This paper proposes the Stay-Move tree (SM tree) to represent frequency distributions for types of trajectories by introducing concepts of stay and move. The proposed tree model was applied to analyzing the Korean Household Travel Survey data. The preliminary results show that the proposed SM trees can potentially be employed to compare/classify spatiotemporal trajectories of different groups (e.g., demographic groups or species of animals). The methodology can potentially be useful to summarize big trajectory data observed from both human and natural phenomena.
Original language | English |
---|---|
Title of host publication | Spatial Big Data and Machine Learning in GIScience, Workshop at GIScience 2018 |
Editors | Martin Raubal, Shaowen Wang, Mengyu Guo, David Jonietz, Peter Kiefer |
DOIs | |
Publication status | Published - 28 Aug 2018 |
Externally published | Yes |