Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.
map generalization is not fully automated and requires important human intervention,
this second tier is necessary to store the results. The third tier is comprised of
the datamarts, which can be further processed and organized according to a vertical
view of the data (e.g., within a range of map resolutions) or a horizontal view (e.g.,
within a region or a department). The fourth tier includes SOLAP clients that can
add local information. Such architecture is useful when the fusion of detailed source
data represents important efforts that cannot be fully automated. Bernier and B?©dard
(2005) describe the difficulties related to spatial data warehousing.
The SOLAP concepts support the multidimensional paradigm and enriched data
exploration based on an explicit spatial reference represented on maps. This explicit
spatial reference can relate to dimensions and measures as SOLAP supports ???spatial???
dimensions and ???spatial??? measures. Three types of spatial dimensions can be
defined: the nongeometric spatial dimensions, the geometric spatial dimensions, and
the mixed spatial dimensions (B?©dard, Merrett, & Han, 2001). In the first type of
spatial dimension, the spatial reference uses nominal data only (e.g., place names)
as no geometry or cartographic representation is associated with the dimension
members.
Pages:
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566