A Class of Region-preserving Space Transformations for Indexing High-dimensional Data | Science Publications

Journal of Computer Science

Jack Lukaszuk¹  Ratko Orlandic¹ 

关键词: Database Systems;    Access Methods;    Space-partitioning Strategy;    Data Dimensionality;   

DOI  :  10.3844/jcssp.2005.89.97

学科分类：计算机科学（综合）

来源: Science Publications

【 摘 要 】

This study introduces a class of region preserving space transformation (RPST) schemes foraccessing high-dimensional data. The access methods in this class differ with respect to their space-partitioningstrategies. The study develops two new static partitioning schemes that can split eachdimension of the space within linear space complexity. They also support an effective mechanism forhandling skewed data in heavily sparse spaces. The techniques are experimentally compared to thePyramid Technique, which is another example of static partitioning designed for high-dimensionaldata. On real high-dimensional data, the proposed RPST schemes outperform the Pyramid Techniqueby a significant margin.

【 授权许可】

Unknown