This chapter performs an analysis to determine the factors that affect the performance of predictive queries and shows that several of these factors are not considered by the Time Parameterized R-tree (TPR-tree), which uses the insertion/deletion algorithms of the R-tree designed for static data. A predictive spatio-temporal query retrieves the set of moving objects that will intersect a query window during a future time interval. The chapter proposes a new index structure called the TPR*-tree, which takes into account the unique features of dynamic objects through a set of improved construction algorithms. The TPR*-tree improves the TPR-tree by employing a new set of insertion and deletion algorithms. In addition, it provides cost models that determine the optimal performance achievable by any data-partition spatio-temporal access method. Using experimental comparison, it illustrates that the TPR-tree is nearly-optimal and significantly outperforms the TPR*-tree under all conditions.
|Original language||English (US)|
|Title of host publication||Proceedings 2003 VLDB Conference|
|Subtitle of host publication||29th International Conference on Very Large Databases (VLDB)|
|Number of pages||12|
|State||Published - Jan 1 2003|
ASJC Scopus subject areas
- Computer Science(all)