Dimensionality reduction for dimension-specific search.

Abstract

Dimensionality reduction plays an important role in efficient similarity search, which is often based on k-nearest neighbor (k-NN) queries over a high-dimensional feature space. In this paper, we introduce a novel type of k-NN query, namely conditional k-NN (ck-NN), which considers dimension-specific constraint in addition to the inter-point distances. However, existing dimensionality reduction methods are not applicable to this new type of queries. We propose a novel Mean-Std(standard deviation) guided Dimensionality Reduction (MSDR) to support a pruning based efficient ck-NN query processing strategy. Our preliminary experimental results on 3D protein structure data demonstrate that the MSDR method is promising.

Citation

HUANG, Z., SHEN, H., ZHOU, X., SONG, D. and RUGER, S. 2007. Dimensionality reduction for dimension-specific search. In Proceedings of the 30th Annual international Association of Computing Machinery Special Interest Group on Information Retrieval (ACM SIGIR) conference on research and development in information retrieval (SIGIR'07), 23-27 July 2007, Amsterdam, Netherlands. New York: ACM [online], pages 849-850. Available from: https://doi.org/10.1145/1277741.1277940

• Presentation Conference Type: Poster
• Conference Name: 30th Annual international Association of Computing Machinery Special Interest Group on Information Retrieval (ACM SIGIR) conference on research and development in information retrieval (SIGIR'07)
• Conference Location: Amsterdam, Netherlands
• Start Date: Jul 23, 2007
• End Date: Jul 27, 2007
• Publication Date: Nov 30, 2007
• Deposit Date: May 13, 2009
• Publicly Available Date: May 13, 2009
• Publisher: ACM Association for Computing Machinery
• Pages: 849-850
• DOI: https://doi.org/10.1145/1277741.1277940
• Keywords: Algorithms; Retrieval models; Scientific databases
• Public URL: http://hdl.handle.net/10059/338