Date of Award

8-2006

Degree Type

Thesis

Degree Name

M.S.

Degree Program

Computer Science

Department

Computer Science

Major Professor

Abdelguerfi, Mahdi

Second Advisor

Tu, Shengru

Third Advisor

Chaudry, Nauman

Abstract

Spatial network queries, for example KNN or range, operate on systems where objects are constrained to locations on a network. Current spatial network query algorithms rely on forms of network traversal which have a high complexity proportional to the size of the network making, them poor for large real-world networks. In this thesis, an alternative method of approximating the results of spatial network queries with a high level of accuracy is introduced. Distances between network points are stored in an M-Tree index, a balanced tree index where metric distance determines data ordering. The M-Tree uses the chessboard metric on network points embedded in a higher dimensional space using tRNE. Using the M-Tree both KNN and range queries are computed more efficiently than network traversal. Error rates of the M-Tree are low, with accuracies of 97% possible on KNN queries and perfect accuracy with 2% extra results on range queries.

Rights

The University of New Orleans and its agents retain the non-exclusive license to archive and make accessible this dissertation or thesis in whole or in part in all forms of media, now or hereafter known. The author retains all other ownership rights to the copyright of the thesis or dissertation.

Download

Share

COinS