Internship Programmes
 
Efficient Mining of Spatio-temporal Data to Discover Trends or Patterns

Discipline: Computer science, computational science, data mining, algorithms

Project Description:
The project aims to study how to find interesting information from spatio-temporal data efficiently. Given a set of location data (in 2D) of some moving objects over time, how can we efficiently find relationships among different objects? In addition, if a group of objects occupy a certain area, is there a way to quantify their effective coverage of the area?

The challenge of the project:
The solution the problem may be computationally intensive. How then can this be done efficiently? For the coverage problem, what is a good way to quantify the coverage for a group of objects so we can determine what is a good coverage. How do we quantify this as the coverage change over time?

Pre-requisite:
Interest in algorithm development, basic programming skill in Java