TY - GEN
T1 - Materialization trade-offs in hierarchical shortest path algorithms
AU - Shekhar, Shashi
AU - Fetterer, Andrew
AU - Goyal, Brajesh
PY - 1997
Y1 - 1997
N2 - Materialization and hierarchical routing algorithms axe becoming important tools in querying databases for the shortest paths in time-critical applications like Intelligent Transportation Systems (ITS), due to the growing size of their spatial graph databases [16]. A hierarchical routing algorithm decomposes the original graph into a set of fragment graphs and a boundary graph which summarizes the fragment graphs. A fully materialized hierarchical routing algorithm pre-computes and stores the shortest-path view and the shortest-path-cost view for the graph fragments as well as for the boundary graph [9]. The storage cost of the fully materialized approach can be reduced by a virtual or a hybrid materialization approach, where few or none of the relevant views are pre-computed. This paper explores the effect of materializing individual views for the storage overhead and computation time of hierarchical routing algorithms. Our experiments with the Twin Cities metropolitan road-map show that materializing the shortest-path-cost view for the boundary graph provides the best savings in computation time, for a given amount of storage and a small number of fragments. Materializing the relevant part of the shortest-path-cost view for the fragment graphs provides the next best savings, followed by materializing the shortest-path view for the boundary graph. Virtual shortest-path-view on fragments can reduce storage costs by an order of magnitude or more for large graphs.
AB - Materialization and hierarchical routing algorithms axe becoming important tools in querying databases for the shortest paths in time-critical applications like Intelligent Transportation Systems (ITS), due to the growing size of their spatial graph databases [16]. A hierarchical routing algorithm decomposes the original graph into a set of fragment graphs and a boundary graph which summarizes the fragment graphs. A fully materialized hierarchical routing algorithm pre-computes and stores the shortest-path view and the shortest-path-cost view for the graph fragments as well as for the boundary graph [9]. The storage cost of the fully materialized approach can be reduced by a virtual or a hybrid materialization approach, where few or none of the relevant views are pre-computed. This paper explores the effect of materializing individual views for the storage overhead and computation time of hierarchical routing algorithms. Our experiments with the Twin Cities metropolitan road-map show that materializing the shortest-path-cost view for the boundary graph provides the best savings in computation time, for a given amount of storage and a small number of fragments. Materializing the relevant part of the shortest-path-cost view for the fragment graphs provides the next best savings, followed by materializing the shortest-path view for the boundary graph. Virtual shortest-path-view on fragments can reduce storage costs by an order of magnitude or more for large graphs.
UR - http://www.scopus.com/inward/record.url?scp=84885602912&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84885602912&partnerID=8YFLogxK
U2 - 10.1007/3-540-63238-7_26
DO - 10.1007/3-540-63238-7_26
M3 - Conference contribution
AN - SCOPUS:84885602912
SN - 3540632387
SN - 9783540632382
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 94
EP - 111
BT - Advances in Spatial Databases - 5th International Symposium, SSD 1997, Proceedings
T2 - 5th International Symposium on Spatial Databases, SSD 1997
Y2 - 15 July 1997 through 18 July 1997
ER -