In Data cube [GBA96,RS97,AAD96,ZDN97,HRU96] , aggregates compute along multiple dimension and
granularities. These are very useful in OLAP or multidimensional data analysis
applications. For example an application could perform an
aggregate computation along the time dimension, say every year, every season
and so on. A straightforward way to compute such aggregates is by grouping for
every subset, but is not viable due to bad performance. User-defined
aggregates allow the user to define such aggregates in a much simpler and
efficient manner.