In the context of hierarchical clustering,
star query processing changes radically. In this chapter, the authors present a
complete abstract processing plan that captures all the necessary steps in evaluating
star queries over hierarchically clustered fact tables. Furthermore, the authors discuss
issues on optimizing star queries within the context of the abstract processing
plan and they define the abstract operations in terms of physical operations over
the CUBE File data structure.
Chapter VII, Bitmap Indices for Data Warehouses, by Kurt Stockinger and Kesheng
Wu, overviews the issues related to a special kind of index used for optimizing
OLAP queries, namely, the bitmap index. Typically, bitmap indexes work well for
attributes of low cardinality since the indexes are small for such atrributes. The
higher cardinality of an indexed attribute, the larger size of a bitmap index. In order
to reduce the sizes of bitmap indexes various techniques are used. This chapter overviews
such techniques, namely, encoding, compression, and binning and it focuses
on a particular compression technique called a word-aligned-hybrid compression.
xvii
Moreover, the authors present multiple experimental results comparing different
encoding techniques and showing the characteristics of the word-aligned-hybrid
compression. The results indicate that for high cardinality attributes compressed
bitmap indexes also offer good index characteristics with respect to their sizes and
query response times.
Pages:
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37