InterSystems Corp??™s
Cache also has bitmap index support since version 5.0.
Even though we do not have technical details on most of these commercial products,
it is generally clear that they tend to use either the basic bitmap index or the
bit-sliced index. Strategies like binning and multicomponent encoding are not used
partly because there is no robust strategy to select parameters like the number of
bins or the number of components that suits different applications.
Summary and Open Problems
In this chapter, we reviewed a number of recent developments in the area of bitmap
indexing technology. We organized much of the research work under the three orthogonal
categories of encoding, compression, and binning. We also provided a brief
overview of commercial bitmap index implementations by major vendors.
Most of the indexing methods reviewed were designed to efficiently answer multidimensional
range queries. However, they are also efficient for other types of
queries, such as joins on foreign keys and computations of aggregates (O??™Neil &
Quass, 1997).
Despite the success of bitmap indices, there are a number of important questions that
remain to be addressed. For example, is there an efficient bitmap index for similar-
6 Stockinger & Wu
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of
Idea Group Inc.
Pages:
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345