Database Reference
In-Depth Information
of a trajectory changes. The minimum description length (MDL) principle is
adopted in this process.
2. Grouping: using given distance measure trajectory segments that are close to each
other are grouped into a cluster. Density-based clustering algorithm is used in this
process, which allows clusters in TRACULAS have any size and shape.
3. Representing: derive a representative trajectory for each cluster. The purpose of
this representative trajectory is to describe the overall movement of the trajectory
partitions that belong to the cluster.
An important step in TRACLUS is to partition a trajectory into sub-trajectories. By
clustering sub-trajectories instead of the whole trajectories, we are able to discover
the common paths shared by different sub-trajectories.
6
Summary
This chapter discusses many interesting state-of-the-art methods of spatiotemporal
pattern mining. Discovery of spatiotemporal patterns can benefit various appli-
cations, such as ecological studies, traffic planning and social network analysis.
We categorize the patterns as individual periodic patterns, pairwise patterns, and
aggregate patterns over multiple trajectories.
As the collection of spatiotemporal data becomes easier and popular, spatiotempo-
ral data mining is a promising research area with a lot of potential interesting research
topics. There are still many challenging issues have not been well addressed by cur-
rent methods, such as sparsity, uncertainties and noises in the data. It is also important
to consider the spatial semantics (e.g., point of interest information) and constraints
(e.g., road network and landscapes). So we could better understand the semantic
meanings of the patterns. Finally, it will be interesting to consider human factor in
the mining process and make the mining process more interactive and informative.
References
1. R. Agrawal and R. Srikant. Mining sequential patterns. In
Proc. 1995 Int. Conf. Data
Engineering (ICDE'95)
, pages 3-14. IEEE, 1995.
2. M. Andersson, J. Gudmundsson, P. Laube, and T. Wolle. Reporting leaders and followers
among trajectories of moving point objects.
GeoInformatica
, 12(4):497-528, 2008.
3. H. Cao, N. Mamoulis, and D. W. Cheung. Discovery of periodic patterns in spatiotemporal
sequences.
Knowledge and Data Engineering, IEEE Transactions on
, 19(4):453-467, 2007.
4. L. Chen and R. T. Ng. On the marriage of lp-norms and edit distance. In
Proc. 2004 Int. Conf.
Very Large Data Bases (VLDB'04)
, 2004.
5. L. Chen, M. T. Özsu, and V. Oria. Robust and fast similarity search for moving object
trajectories. In
Proc. 2005 ACM-SIGMOD Int. Conf. Management of Data (SIGMOD'05)
,
2005.
6. J. Cranshaw, E. Toch, J. I. Hong, A. Kittur, and N. Sadeh. Bridging the gap between phys-
ical location and online social networks. In
Proc. 2010 Int. Conf. Ubiquitous Computing
(Ubicomp'10)
, 2010.
