In many database applications similarity search is a typical operation that must be efficiently processed. A class of techniques to implement similarity queries which has demonstrated acceptable results includes the use of a nearest neighbor search algorithm on top of a multidimensional access method. We propose a newalgorithm based on the space organization induced by the Q-tree. Specifically, a metric is used to discard internal zones of a cube that have been previously extracted from it. Our experiments based on both synthetic and real data sets showthat our approach outperforms other competitive techniques like R*-tree, M-tree, and Hybrid-tree. © Springer-Verlag Berlin Heidelberg 2002.
CITATION STYLE
Jurado, E., & Barrena, M. (2002). Efficient similarity search in feature spaces with the Q-Tree. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2435 LNCS, pp. 177–190). Springer Verlag. https://doi.org/10.1007/3-540-45710-0_15
Mendeley helps you to discover research relevant for your work.