C. Lang, A. K. Singh. Modeling High-Dimensional Index Structures using Sampling. In ACM SIGMOD Conference, Santa Barbara, CA, 2001. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Indexing without the Index: Scalable.. - Riedewald, Agrawal, .. (2002) (Correct)
....of a query. However, it is specific to our technique and can take advantage of the regular and simple structure of MFS. For data structures like B trees and R trees the shape of the bounding box of a tree node can only be approximated which leads to high errors with increasing dimensionality [19]. The well known data cube [11] is a d dimensional generalization of the GROUP BY operator. It computes all 2 possible results of grouping the d dimensional base data set by any subset of its dimensions. MFS provides a new way of supporting range aggregate queries for one or more of the ....
C. A. Lang and A. K. Singh. Modeling highdimensional index structures using sampling. In of Data, pages 389--400, 2001.
Efficient Nearest Neighbor Retrieval by Using a Local.. - Balko, Schmitt (2002) (Correct)
....software components (e.g. operating system swap cycles) Throughout this section we use some notational conventions. Unless explicitly stated di erently, the distinct experimental results appear in the diagrams with the shapes from Table 4.1. Furthermore, we use the sampling method proposed in [LS01] for our experiments. In this way, we ran 1000 queries with distinct query points for each measurement reading and computed an average value. The retrieval testbed was implemented in Java interacting with an Oracle database on top of a Linux workstation. UNIFORM COREL, ADAC AV tree H N VA ....
C. A. Lang and A. K. Singh. Modeling High-Dimensional Index Structures using Sampling. In Walid G. Aref, editor, Proc. of the
Accelerating High-dimensional Nearest Neighbor Queries - Lang, Singh (2002) Self-citation (Lang Singh) (Correct)
....their high accuracy in modeling data by considering local effects. A disadvantage of the histogram approaches is that they are not applicable in high dimensions since either the number of histogram regions becomes too large or these regions contain too much empty space and become inaccurate. In [22], sampling is used to overcome this problem. In contrast to this paper, the sample is used to predict the overall query cost of a given index structure. More specifically, the sample is used to predict the index page layout. Here, on the other hand, we use the sample to predict the query radius ....
....query cost of a given index structure. More specifically, the sample is used to predict the index page layout. Here, on the other hand, we use the sample to predict the query radius for one particular query. Moreover, we employ the fractal dimensionality to compensate for sampling, while in [22], uniformity is assumed to adjust the page layout prediction. Accelerated All NN Query Algorithms Several papers (e.g. 11, 16] focus on efficient processing of spatial joins on R trees. However, they do not take seek and transfer costs into account. Instead, they try to minimize the number of ....
Christian A. Lang and Ambuj K. Singh. Modeling highdimensional index structures using sampling. In Proc. 2001.
A Framework for Accelerating High-dimensional NN-queries - Lang, Singh (2001) (2 citations) Self-citation (Lang Singh) (Correct)
....their high accuracy in modeling data by considering local effects. A disadvantage of the histogram approaches is that they are not applicable in high dimensions since either the number of histogram regions becomes too large or these regions contain too much empty space and become inaccurate. In [16], sampling is used to overcome this problem. In contrast to this paper, the sample is used to predict the overall query cost of a given index structure. More specifically, the sample is used to predict the index page layout. Here, on the other hand, we use the sample to predict the query radius ....
....query cost of a given index structure. More specifically, the sample is used to predict the index page layout. Here, on the other hand, we use the sample to predict the query radius for one particular query. Moreover, we employ the fractal dimensionality to compensate for sampling, while in [16], uniformity is assumed to adjust the page layout prediction. We are not aware of any framework that transforms NN queries into at most two range queries and is thereby able to speed up the query performance of any given index structure. 6 Conclusions We showed in this paper that it is possible ....
Christian A. Lang and Ambuj K. Singh. Modeling high-dimensional index structures using sampling. In Proc. ACM SIGMOD Int. Conf. on Management of Data, 2001.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC