基于查询采样的高维数据混合索引-CiteSeerX.PDF

ISSN 1000-9825, CODEN RUXUEW E-mail: jos@ Journal of Software , Vol.19, No.8, August 2008, pp.2054−2065 DOI: 10.3724/SP.J.1001.2008.02054 Tel/Fax: +86-10 © 2008 by Journal of Software . All rights reserved. ∗ 基于查询采样的高维数据混合索引 1,2,3 3+ 3 张军旗 , 周向东 , 施伯乐 1(北京大学 信息科学技术学院 智能科学系,北京 100871) 2(北京大学 信息科学技术学院 机器感知与智能教育部重点实验室,北京 100871) 3(复旦大学 计算机与信息技术系,上海 200433) High Dimensional Hybrid Index Based on Query Sampling ZHANG Jun-Qi1,2,3, ZHOU Xiang-Dong3+, SHI Bai-Le3 1(Department of Machine Intelligence, School of Electronics Engineering and Computer Science, Peking University, Beijing 100871) 2(Key Laboratory of Machine Perception for the Ministry of Education, School of Electronics Engineering and Computer Science, Peking University, Beijing 100871) 3(Department of Computing and Information Technology, Fudan University, Shanghai 200433) + Corresponding author: E-mail: xdzhou@, Zhang JQ, Zhou XD, Shi BL. High dimensional hybrid index based on query sampling. Journal of Software, 2008,19(8):2054−2065. /1000-9825/19/2054.htm Abstract : In order to improve the query answering of high-dimensional database, data distribution is necessary to select appropriate indexing strategy. However, traditional data distribution models can not estimate the accurate data distribution in the complex real multimedia data of image and video. This paper presents a method to estimate the accurate data distribution based on query sampling, and proposes a novel hybrid index to