For all query dimensions the range-encoded bitmap index with
100 bins shows the best performance characteristics, however, sometimes at the cost
of a larger index. In case the storage space is a limiting factor, it is better to choose
equality-encoded bitmap indices with 1000 bins (see Figure 7). As we can see in
Figure 9, the performance of equality-encoded bitmap indices is not significantly
different from the performance of range-encoded bitmap indices.
Key Features in Commercial Products
Due to the considerable amount of work involved in producing and maintaining a
robust commercial software system, only the most efficient and proven indexing
Figure 9. Multidimensional queries with various bitmap indices. EE-1000: equality
encoding with 1000 bins, RE-100: range encoding with 100 bins
0.
0
0. 0. 0.0 0.00 0.000 0.0000
a) Query box size of 2D queries
Time [sec]
Project on Index EE- 000 RE- 00
0.
0
00
0. 0. 0.0 0.00 0.000 0.0000
b) Query box size of 10D queries
Time [sec]
Project on Index EE- 000 RE- 00
Bitmap Indices for Data Warehouses
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.
technologies make their way into a commercial DBMS. In this section, we give
a short review of the key bitmap indexing technologies currently used by various
well-known commercial products.
Pages:
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343