These functional and inclusion dependencies can be combined and
used for the optimization of the grouping and join operations. The complex optimization
technique that exploits these existing integrity constraints is the hierarchical
pregrouping and it is presented next. Experimental results have shown that this
technique can drastically reduce the execution time of the examined OLAP queries
(Pieringer et al., 2003).
Other optimization opportunities exist and pertain to the Create_Range operation
of the abstract processing plan (Figure 5(a)), or the exploitation of the sort-order of
the tuples coming from the fact table (MD_Range_Access operation). Due to lack
of space, these techniques will not be described here. The interested reader can
find details in Karayannidis et al. (2002), Tsois and Sellis (2003), Pieringer et al.
(2003), and Tsois (2005) for the former and in Theodoratos and Tsois (2003) and
Tsois (2005) for the latter.
The Hierarchical Pregrouping technique is based on the properties of the join and
grouping operations. The grouping operation uses the values of the grouping attributes
only to group tuples that have the same value. However, the actual value of
a grouping attribute is not important. Therefore, an attribute X that is used only in
a grouping operation, like Group_Select, can be replaced with any other attribute
Y when there is a bijective (1-1 and onto) mapping among the values of X and Y.
Pages:
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302