Theoretical (Analytical):

Practical (Implementation):

Literature Work:

Overview

Spatiotemporal clustering is a process of grouping objects based on proximity within the spatial and temporal component.

With this technique, large datasets can be summarized and aggregated to permit a more practical analysis and provide meaningful insights.

 

Problem Statement

Due to the possible volativity in both the spatial and temporal component, clusters may

  • split or merge
  • change their location
  • exist only at specific times
  • change in shape or size

and thus alter their intrinsic characteristics.

The effective visualization of clusters in spatiotemporal sequences, especially in large datasets, thus remains a challenging endeavour.

 

Tasks

  • Evaluate and compare existing approaches for visual cluster detection in spatiotemporal datasets
  • Develop a novel interactive, visual spatiotemporal clustering technique
  • Compare approach to existing spatiotemporal clustering algorithms

Requirements

  • Interest in spatiotemporal data
  • Basic knowledge about information visualization, data mining and geographic information systems
  • Good programming skills in: JavaScript/TypeScript, D3.js, Python

 

Scope/Duration/Start

  • Scope: Bachelor / Master
  • 3 Month Project, 3 / 6 Month Thesis
  • Start: immediately

Contact

References

  • Bach, B., A Descriptive Framework for Temporal Data Visualizations Based on Generalized Space-Time Cubes. Computer Graphics Forum, 36 (2017). https://doi.org/10.1111/cgf.12804
  • Andrienko, G., Andrienko, A., Interactive cluster analysis of diverse types of spatiotemporal data. SIGKDD Explor. Newsl. 11, 2 (2009). https://doi.org/10.1145/1809400.1809405
  • Chen, W., Huang, Z., Wu, F., Zhy, M., Guan, H., Maciejewski, R. , A Visual Analysis Approach for Exploring Spatio-Temporal Urban Data. IEEE Trans Vis Comput Graph. (2018). https://doi.org/10.1109/TVCG.2017.2758362