Databases Reference
In-Depth Information
Figure 2.5
Bitmaps and query processing.
attributes or salaried/hourly employee attributes. Bitmap indexes speed up special index
operations such as intersection or union. In a typical B+tree, secondary index, for
instance, leaf nodes have a pointer array of physical row addresses for all rows that meet
a given criterion defined by the index.
In a bitmap, however, the structure is different. Each row a table is represented by a
single bit in a bit vector, and each row satisfying the criterion (e.g., reg-name = 'south-
west' or sales-id = 412) has its bit set in that bit vector, as shown in Figure 2.5. Here we
see that each attribute value has a separate bit vector. Those rows satisfying all criteria
