g., dimension of Figure 2) but
also by the mixture of hierarchies in the dimension (e.g., dimension of Figure 4).
Adapting. Heterogenous. Dimensions
The general approach to handle structural heterogeneity in OLAP is to adapt or
transform the dimensions in order to obtain homogeneous data. In this section we
examine two approaches along this idea. First, we explain the use of null elements,
and then we explore structural adjustments of heterogeneous dimension necessary
to obtain homogeneous dimensions. In the latter case, the hierarchy schemas of the
original dimensions are modified.
Table 2.
Brand Category
b1 c1
b1 c2
b2 c1
b2 c2
b3 c2
Handling Structural Heterogeneity in OLAP 4
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.
Null Elements
At first sight, as heterogeneity is nothing more than having elements without parents
in some categories, the problem could be solved straightforwardly by adding null
elements. In this form a heterogeneous dimension would be turned homogeneous
without any modification of its structure. However, it is not easy to add null elements
to get a homogeneous dimension whose rollup relations are strict, that is, without
violating the fundamental property of dimensions explained. Using a single null
element null, as in the relational database setting, does not prevent this problem.
Pages:
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115