Memory-adaptative dynamic spatial approximation trees

2002

The Spatial Approximation Tree (sa-tree) is a recently proposed data structure for searching in metric spaces. It has been shown that it compares favorably against alternative data structures in spaces of high dimension or queries with low selectivity. The main drawback of the ...

2010

The metric space model allows abstracting many similarity search problems. Similarity search has multiple applications especially in the multimedia databases area. The idea is to index the database so as to accelerate similarity queries. Although there are several promising indices, few of them are dynamic, i.e., once created very few allow to perform insertions and deletions of elements at a reasonable cost.

Similarity Search and Applications, 2009 …, 2009

Metric space searching is an emerging technique to address the problem of efficient similarity searching in many applications, including multimedia databases and other repositories handling complex objects. Although promising, the metric space approach is still immature in several aspects that are well established in traditional databases. In particular, most indexing schemes are not dynamic, that is, few of them tolerate insertion of elements at reasonable cost over an existing index and only a few work efficiently in secondary memory.

Journal of Computer Science Technology, 2014

Metric space searching is an emerging technique to address the problem of similarity searching in many applications. In order to efficiently answer similarity queries, the database must be indexed. In some interesting real applications dynamism is an indispensable property of the index. There are very few actually dynamic indexes that support not only searches, but also insertions and deletions of elements. The dynamic spatial approximation tree (DSAT) is a data structure specially designed for searching in metric spaces, which compares favorably against other data structures in high dimensional spaces or queries with low selectivity. Insertions are efficient and easily supported in DSAT, but deletions degrade the structure over time. Several methods are proposed to handle deletions over the DSAT. One of them has shown to be superior to the others, in the sense that it permits controlling the expected deletion cost as a proportion of the insertion cost and searches does not overly degrade after several deletions. In this paper we propose and study a new alternative deletion method, based on the better existing strategy. The outcome is a fully dynamic data structure that can be managed through insertions and deletions over arbitrarily long periods of time without any significant reorganization.

Chilean Computer Science Society, …, 2003

The Dynamic Spatial Approximation Tree (dsa-tree) is a recently proposed data structure for searching in metric spaces. It has been shown that it compares favorably against alternative data structures in spaces of high dimension or queries with low selectivity. The dsa-tree supports insertion and deletions of elements. However, it has been noted that deletions degrade the structure over time, so the structure cannot be regarded as fully dynamic in the sense that deletions are not sustainable for long periods of time.

2010

Metric space searching is an emerging technique to address the problem of efficient similarity searching in many applications, including multimedia databases and other repositories handling complex objects. Although promising, the metric space approach is still immature in several aspects that are well established in traditional databases. In particular, most indexing schemes are not dynamic. From the few dynamic indexes, even fewer work well in secondary memory. That is, most of them need the index in main memory in order to operate efficiently. In this paper we introduce a secondary-memory variant of the Dynamic Spatial Approximation Tree with Clusters (DSACL-tree) which has shown to be competitive in main memory. The resulting index handles well the secondary memory scenario and is competitive with the state of the art. The resulting index is a much more practical data structure that can be useful in a wide range of database applications.

2008

Many computational applications need to look for informa- tion in a database. Nowadays, the predominance of non- conventional databases makes the similarity search (i.e., searching elements of the database that are "similar" to a given query) becomes a preponderant concept. The Spatial Approximation Tree has been shown that it compares favorably against alternative data structures for similarity searching in metric spaces of medium to high di- mensionality ("difficult" spaces) or queries with low selec- tivity. However, for the construction process the tree root has been randomly selected and the tree ,in its shape and performance, is completely determined by this selection. Therefore, we are interested in improve mainly the searches in this data structure trying to select the tree root so to re- flect some of the own characteristics of the metric space to be indexed. We regard that selecting the root in this way it allows a better adaption of the data structure ...

2013

The Dynamic Spatial Approximation Tree (DSAT) is a data structure specially designed for searching in metric spaces. It has been shown that it compares favorably against alternative data structures in spaces of high dimension or queries with low selectivity. The DSAT supports insertion and deletions of elements. However, it has been noted that eliminations degrade the structure over time. In [8] is proposed a method to handle deletions over the DSAT, which shown to be superior to the former in the sense that it permits controlling the expected deletion cost as a proportion of the insertion cost. In this paper we propose and study a new deletion method, based on the deletions strategies presented in [8], which has demonstrated to be better. The outcome is a fully dynamic data structure that can be managed through insertions and deletions over arbitrarily long periods of time without any reorganization.

Journal of Information and Data Management

Many applications rely on spatial information retrieval, which involves costly computational geometric algorithms to process spatial queries. Spatial approximations simplify the geometric shape of complex spatial objects, allowing faster spatial queries at the expense of result accuracy. In this sense, spatial approximations have been employed to efficiently reduce the number of objects under consideration, followed by a refinement step to restore accuracy. For instance, spatial index structures employ spatial approximations to organize spatial objects in hierarchical structures (e.g., the R-tree). It leads to the interest in studying how spatial approximations can be efficiently employed to improve spatial query processing. This article presents a systematic review on this topic. We gather relevant studies by performing a search string on several digital libraries. We further expand the studies under consideration by employing a single iteration of the snowballing approach, where w...

2009

Mobile query processing is, currently, a very active research field. Range and nearest neighbor queries are commonly used in spatiotemporal databases and location based services (LBS). In this paper, we focus on finding nearest neighbors of a query point within a certain distance range. We propose a new indexing structure CN-tree, Compact N-tree, based on a recent indexing technique called N-tree. CN-tree joins efficiency of N-tree's data partitioning scheme to pertinent objects' approximation with minimal bounding rectangles of R-trees which are reported to be the best performing for range search. We show how we use the approximation in constructing CN-tree and, then, how this index can support range queries efficiently by minimizing computation of distances and avoiding overlapping of minimal bounding rectangles. The experimental results through the comparison with the well know R*-tree, show that the proposed CN-tree widely outperforms R*-tree as an in-memory index and it presents competitive performances when used as an in-disk index.