Data mining techniques emerged as a result of long processes of development and
evolution in areas such as statistics, artificial intelligence and machine learning. They have
not only pointed to quantifiable business benefits, but also made a qualitative revolutionary
step forward in the sense that they enabled new ways of formulating business queries,
where answers are obtained accurately and quickly. The innovativeness of these tools is
reflected primarily in the radical turn away from the retrospective data access that used to be
typical of decision support systems, toward prospective and proactive information delivery.
The old business questions of the type ‘‘What was our revenue in the past period?’’, which
were typical of collection and storage of data in the period from the 1960s to 1990s, have
changed into a new pattern of questions saying: ‘‘What will our revenue be in the coming
period and why‘‘ (Dillon, 1998; Cahlink, 2000; Han and Kamber, 2001)?