A revised r*-tree in comparison with related index structures

  • Authors:

  • Norbert Beckmann;Bernhard Seeger

  • Affiliations:

  • University of Bremen, Bremen, Germany;University of Marburg, Marburg, Germany

  • Venue:
  • Proceedings of the 2009 ACM SIGMOD International Conference on Management of data
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper we present an improved redesign of the R*-tree that is entirely suitable for running within a DBMS. Most importantly, an insertion is guaranteed to be restricted to a single path because re-insertion could be abandoned. We re-engineered both, subtree choice and split algorithm, to be more robust against specific data distributions and insertion orders, as well as peculiarities often found in real multidimensional data sets. This comes along with a substantial reduction in CPU-time. Our experimental setup covers a wide range of different artificial and real data sets. The experimental comparison shows that the search performance of our revised R*-tree is superior to that of its three most important competitors. In comparison to its predecessor, the original R*-tree, the creation of a tree is substantially faster, while the I/O cost required for processing queries is improved by more than 30% on average for two- and three-dimensional data. For higher dimensional data, particularly for real data sets, much larger improvements are achieved.