C. B ohm and H.-P. Kriegel. A cost model and index architecture for the similarity join. In Proc. of ICDE, pages 411--420, 2001.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....the trees in depth first order, expanding children of a node pair if these nodes intersect. Further optimizations are achieved by using either local plane sweep order, pinning, or local z order. The performance of similarity joins can also be improved by decoupling CPU and I O optimizations [7]. BFRJ [23] tra verses the R tree in Breadth First Search order. The authors show that this ordering reduces the number of accesses to the pages of the R tree. Lo and Ravishankar [31] consider the spatial join problem when the R tree is precomputed on only one of the data sets. They propose to ....

C. B6hm and H.P. Kriegel. A cost model and index architecture for the similarity join. In ICDE, Heidelberg, Germany, 2001.

....8 discusses cluster ordering. Section 9 presents the experimental results. We end with a brief discussion in Section 10. 2 Related work Joining two datasets is a costly operation. Current techniques reduce this cost by pruning pairs of data Without Index With Index point data [6, 7, 19, 44] [8, 11, 24] spatial data [3, 12, 29, 30, 36, 38, 45] 5, 20, 23, 31, 32] Table 1. A classification of join techniques. points that do not appear in the final join. They can be classified into two groups based on the data structures they use: 1) no index is built on the datasets, or 2) index is built only ....

....the expansion to only those children that intersect with the intersection of the parent nodes. Further optimizations are achieved by using either local plane sweep order, pinning, and local z order. The performance of similarity joins can also be improved by decoupling CPU and I O optimizations [8]. This is be done by using an index structure with coarse granularity along with a secondary index structure. BFRJ [24] traverses the R tree in Breadth First Search order. The authors show that this ordering reduces the number of accesses to the pages of the R tree. Lo and Ravishankar [31] ....

C. Bohm and H.P. Kriegel. A cost model and index architecture for the similarity join. In ICDE, Heidelberg, Germany, 2001.

....work in the area of similarity join algorithms. An algorithm based on a new kind of sort order, the epsilon grid order [BBKK 01] has been accepted at the ACM SIGMOD Int. Conf on Management of data. Several further articles regard cost modelling and optimization of the similarity join: In [BK 01a] we propose a cost model for the index based evaluation of the similarity join. The cost model accurately estimates the probability with which a pair of pages must be formed in the joining process (the mating probability) Starting from that point, we optimize the page capacity of an index for ....

Bhm C., Kriegel H.-P.: A Cost Model and Index Architecture for the Similarity Join, IEEE Int. Conf. on Data Engineering, 2001.

....50 speed up factors. Brinkhoff, Kriegel and Seeger adapted RSJ for join processing on parallel computers using shared virtual memory [BKS 96] Their technique improves both CPU time and I O time. Recently, index based similarity join methods have been analyzed from a theoretical point of view. BK 01] proposes a cost model based on the concept of the Minkowski sum [BBKK 97] which can be used for optimizations such as page size optimization. The analysis reveals a serious optimization conflict between CPU and I O. While the CPU requires fine grained partitioning with page capacities of only a ....

....the data set according to the other dimensions until a defined node capacity is reached. For each dimension, the data set is partitioned at most once into stripes of width . Finally, a tree matching algorithm is applied which is restricted to neighboring stripes. It was pointed out in [BK 01] that the kdB tree has very restricting limitations when scaling to large data sets which are not main memory resident. Depending on several parameters such as the data distribution the algorithm needed large portions of the database simultaneously in main memory in order to be operational at ....

[Article contains additional citation context not shown here]

Bhm C., Kriegel H.-P.: A Cost Model and Index Architecture for the Similarity Join, IEEE Int. Conf on Data Engineering (ICDE), 2001.

No context found.

C. B ohm and H.-P. Kriegel. A cost model and index architecture for the similarity join. In Proc. of ICDE, pages 411--420, 2001.

No context found.

C. Bohm and H.-P. Kriegel. A Cost Model and Index Architecture for the Similarity Join. In Proc. IEEE Conf. on Data Engineering, pages 411-420, Heidelberg, Germany, 2001.

No context found.

C. Bohm and H.-P. Kriegel. A cost model and index architecture for the similarity join. In Proceedings of the International Conference on Data Engineering, 2001.

No context found.

C. Bohm and H.-P. Kriegel. A cost model and index architecture for the similarity join. In Proceedings of the International Conference on Data Engineering, 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