Spatial Data Warehouse and Mining

Due to enormous volume of spatial data, it becomes difficult to analysis that volume without having specially designed softwares. Recently, the popularity of spatial information, such as maps created from satellite images and the utilization of telemetry systems, has created repositories of huge amounts of data which need to be efficiently analyzed. Spatial data warehouse is answer to this problem

In analogy to the non-spatial case, a spatial data warehouse can be considered, which supports On Line Analytical Processing (OLAP) operations on both spatial and non-spatial data; this is different from classical data warehouse because data domain is changed.

Various OLAP operations are done on spatial data for decision making and only efficient spatial data warehouse can provide better OLAP operations. However, having an efficient warehouse design itself poses many challenges because of data format.

On the other hand, discovery of spatial association rules may disclose interesting relationships among spatial and/or nonspatial data in large spatial databases and thus it represents a new and promising direction in spatial data mining. Spatial data mining techniques are often derived from spatial statistics and spatial analysis. Other integrals of mining tools are machine learning and database, which are customized to analyze massive data set.

In this project, I aim to this problem by implementing suitable method from the field of Spatial Data Warehouse and Mining. Below report gives a description of the preliminary work done in this direction.

A first stage report can be found here

(this will be keep growing.....)