A Pattern-Aware Prediction Model for Moving Objects
Speaker: Hoyoung Jeung, ITEE
When: 2006-09-15 10:00:00
Venue: 78-420
Host: Dr Heng tao Shen
Abstract:We are facing an unprecedented proliferation of mobile devices, many
equipped with positional technologies such as the GPS. These devices
produce a huge amount of positional data, which needs to be managed
by databases. Due to the large number of applications, predicting
future locations of a moving object has become one of the most
popular research areas. However, existing prediction methods have
some limitations. First, they can not forecast locations accurately
if the query time is far away from the current time. Second, even
for near future prediction, most techniques assume an object follows
some statistical parameters but an object's movements are more
complicated than what their mathematic formulas can represent. In
order to overcome the limitations of existing prediction methods, we
aim to propose a novel prediction model for moving objects, called
the hybrid prediction model, which forecasts future locations based
on trajectory pattern information as well as existing motion
functions. In this approach, we face two main problems. The first
one is how to design trajectory patterns having ad-hoc forms for
prediction and how to discover them from given an object's
trajectory. The second is to consider indexing and query processing
techniques based the trajectory patterns.
Biography:(biography unavailable)
Type: Ph.D confirmation
Contact:Dr Heng tao Shen, seminar host (shenht@itee.uq.edu.au)
or Guido Governatori (ITEE seminar co-ordinator)
(guido@itee.uq.edu.au)