High Impact Factor : 4.396 icon | Submit Manuscript Online icon | UGC Approved icon

Service Management and Allocation with Modified R-Tree Indexing for Spatial Activity Clustering

Author(s):

Midhunlal P V , Mar Athanasius College of Engineering; Jisha P Abraham, Mar Athanasius College of Engineering; Surekha Mariyam Varghese, Mar Athanasius College of Engineering

Keywords:

R-Tree, Spatial Activity Clustering, data mining

Abstract

The process of creating groups from a list of elements based on its properties is called clustering in datamining. Clustering can be applied to the spatial domain in which the clustered elements will be in the spatial network. In Spatial Activity Clustering the elements are the spatial locations of activities occurring in the spatial network. There is a Spatial network and set of activities, the aim is to create k linear paths which are clusters consisting of activities with minimum distances. Its applications is in the area of GIS (Geographic Information Systems) consisting of linear paths such as roadways and observations is in them. Spatial Activity Clustering uses network distance for clustering. Related works are based on geometric distances are less efficient for linear clustering. There is also works using network distance uses subgraph based approaches creates only single routes. The proposed system for Spatial Activity Clustering (SAC) computes shortest paths in spatial network which are the clusters consisting of activities. It uses End Point Joining (EPJ) for finding efficient routes. After the clustering if a need for emergency service arises, service allocation routes for these clusters are created from the service locations which may contain multiple units of service. Accessing spatial data sequentially is a cumbersome task in terms of time and cost. For minimizing the access time, a modified R Tree indexing is used to improve the performance of SAC. This tree creates neighboring links in the leaf nodes based on the spatial proximity. kNN (Nearest Neighbor) Search is used to identify the k nearest summarypaths (cluster) where multiple units can service and service routes are created.

Other Details

Paper ID: SPDM043
Published in: Volume : 1, Issue : 2
Publication Date: 01/12/2015
Page(s): 41-46

Article Preview




Download Article

Share