A dimension table
Advanced Ad Hoc Star Query Processing
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission
of Idea Group Inc. is prohibited.
is organized as a B+ tree with the h-surrogate attribute as the search key. Each tuple
coming from the CUBE File side contains an hsk attribute (i.e., an h-surrogate)
corresponding to the DATE dimension. We use this value as a key for accessing
directly the DATE dimension and retrieving the single tuple that matches. Indeed,
since hsk is a primary key of the dimension table, there will be only a single tuple
match. Therefore, the number of tuples in the output of the IndexResJoin operator
is the same as the one in the input.
Similarly, for each hsk value corresponding to dimension LOCATION we access the
corresponding tuple and retrieve the appropriate Area value. Finally, the grouping
and aggregation has to take place. We depict a hash-group operator that groups the
incoming tuples by Area and Month.
Star Query Optimization
An interesting feature of the database schema is the hierarchical structure of the
h-surrogate attributes and the fact that they encode all the hierarchical attributes
of the dimensions. Therefore, a number of functional dependencies exist in the
schema, along with inclusion dependencies defined through the key-foreign key
relationships.
Pages:
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301