Abstract
In this paper, we discuss the properties of OLAP object sets and range queries and propose pyhisical organizations of object sets together with query processing techniques that attempt both to respect and exploit those properties in answering OLAP partial range queries. Our technique, nicknamed MARS, is a hybrid technique composed of innovative veriants of several well known techniques which can be layered on top of each other. In contrast to other index structures, its modular nature makes it well suited to efficiently support a broad variety of partial range queries. The technique is based on a variant of the extended pyramid tree, columnwise clustering of attribute values, lossless and lossy compression and equi-depth histograms for selectivity estimation. MARS can be implemented both on top of an ODBMS and a RDBMS.
| Original language | English |
|---|---|
| Number of pages | 11 |
| Publication status | Published - Nov 1998 |
Fields of science
- 101004 Biomathematics
- 101027 Dynamical systems
- 101028 Mathematical modelling
- 101029 Mathematical statistics
- 101014 Numerical mathematics
- 101015 Operations research
- 101016 Optimisation
- 101017 Game theory
- 101018 Statistics
- 101019 Stochastics
- 101024 Probability theory
- 101026 Time series analysis
- 102 Computer Sciences
- 102001 Artificial intelligence
- 102003 Image processing
- 102004 Bioinformatics
- 102013 Human-computer interaction
- 102018 Artificial neural networks
- 102019 Machine learning
- 103029 Statistical physics
- 106005 Bioinformatics
- 106007 Biostatistics
- 202017 Embedded systems
- 202035 Robotics
- 202036 Sensor systems
- 202037 Signal processing
- 305901 Computer-aided diagnosis and therapy
- 305905 Medical informatics
- 305907 Medical statistics
- 102032 Computational intelligence
- 102033 Data mining
- 101031 Approximation theory