SEARCH
0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Prev | Current Page 267 | Next

Robert Wrembel and Christian Koncilia

"Data Warehouses and Olap: Concepts, Architectures and Solutions"


Background.
Preliminary.Concepts..
In a relational OLAP (ROLAP) implementation, a dimension is stored into one or
more dimension tables, each having a set of attributes. Dimension attributes usually
form one or more classification hierarchies. For example, the h1 attribute is classified
by the h2 attribute, which is further classified by the h3 attribute, and so forth. We
call the attributes h1, h2, h3, ??¦ hierarchical attributes because they participate in the
definition of the hierarchy. For example, day, month, and year can be a hierarchical
classification in the DATE dimension. For the purposes of this chapter we will assume
a single hierarchy for each dimension.1 A dimension table may also contain
one or more feature attributes f. Feature attributes contain additional information
about a number of hierarchical attributes and are always functionally dependent on
one (or more) hierarchical attribute. For example, population could be a feature attribute
dependent on the region attribute of dimension LOCATION.
Measures (or facts) are stored in fact tables. A fact table may contain one or more
measure attributes and is always linked (by foreign key attributes) to some dimension
tables. This logical organization consisting of a central table (the fact table) and
surrounding tables (the dimension tables) that link to it through 1:N relationships
is known as the star schema (Chaudhuri & Dayal, 1997).


Pages:
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279