The R*-tree: an efficient and robust access method for points and rectangles
SIGMOD '90 Proceedings of the 1990 ACM SIGMOD international conference on Management of data
Data structures and algorithms for nearest neighbor search in general metric spaces
SODA '93 Proceedings of the fourth annual ACM-SIAM Symposium on Discrete algorithms
Indexing large metric spaces for similarity search queries
ACM Transactions on Database Systems (TODS)
Some approaches to best-match file searching
Communications of the ACM
R-trees: a dynamic index structure for spatial searching
SIGMOD '84 Proceedings of the 1984 ACM SIGMOD international conference on Management of data
Fast Indexing and Visualization of Metric Data Sets using Slim-Trees
IEEE Transactions on Knowledge and Data Engineering
M-tree: An Efficient Access Method for Similarity Search in Metric Spaces
VLDB '97 Proceedings of the 23rd International Conference on Very Large Data Bases
Near Neighbor Search in Large Metric Spaces
VLDB '95 Proceedings of the 21th International Conference on Very Large Data Bases
Faster construction of ball-partitioning-based metric access methods
Proceedings of the 28th Annual ACM Symposium on Applied Computing
| Hi-index | 0.00 |
The existing Metric Access Methods (MAM) assume the data elements represent immutable objects. However, many applications must handle complex data evolving over time. Health care, weather monitoring, and other applications require removing or updating elements. Most of the MAM presented in the literature either do not have the deletion operation described, or it is performed just marking the element as deleted without effectively removing it from the structure. In this paper we describe an algorithm that effectively removes any element from a metric tree. While maintaining the height-balancing of the structure, the proposed deletion algorithm uses mechanisms to enforce a reduced number of pages in the tree, improving the query performance. Based on the deletion algorithm, we propose a new way to optimize a MAM, which we call the Push-pulltechnique. It reduces the node overlap performing the deletion and reinsertion of elements close to the border of each node covering region. We also developed the Smart Push-pullalgorithm, which uses statistical data about subtrees' overlapping to calculate how many elements should be removed from each node. The statistics are collected during the evaluation of the structure overlap, an operation employed to ascertain the need to trigger an optimization process. The experiments were run on the Slim-tree and showed a reduction of overlap and a query performance improvement over trees optimized by this technique as compared over trees optimized by the Slim-down method.