The SR-tree: an index structure for high-dimensional nearest neighbor queries
SIGMOD '97 Proceedings of the 1997 ACM SIGMOD international conference on Management of data
ACM Computing Surveys (CSUR)
ACM Transactions on Database Systems (TODS)
Cost models for overlapping and multiversion structures
ACM Transactions on Database Systems (TODS)
Collision Detection for Interactive Graphics Applications
IEEE Transactions on Visualization and Computer Graphics
Cost models for distance joins queries using R-trees
Data & Knowledge Engineering
Bottom-up nearest neighbor search for R-trees
Information Processing Letters
The priority R-tree: A practically efficient and worst-case optimal R-tree
ACM Transactions on Algorithms (TALG)
A new enhancement to the R-tree node splitting
Journal of Information Science
A new double sorting-based node splitting algorithm for R-tree
Programming and Computing Software
Hi-index | 0.00 |
Most of the existing spatial indices are constructed using a single hierarchal index structure; hence a large number of index pages (nodes) are most likely to be inspected during spatial query execution. Since spatial queries usually fetch spatial objects based on their spatial position in the space, it is significant that spatial objects are clustered in such a way that pertinent objects to a query are fetched quickly. This paper presents a method for partitioning the whole space into set of small subspaces. Then, an index structure for each subspace (called the Multi Small Index) is built. This makes it is easy to quickly retrieve spatial objects that are relevant to the query in question using their corresponding small spatial index structures and ignoring other irrelevant indices. To evaluate our new approach, we conducted a set of experimental studies using a collection of real-life spatial datasets (TIGER data files) with diverse sizes and different object sizes, densities and distributions, as well as various query sizes. The results show that (using small query sizes) our proposed structure (Multi Small Index) outperforms the original R-tree (Single Big Index) structure, achieving nearly 50% saving in disk access.