Data Architecture: A Primer for the Data Scientist

3.3

Historical Data

Abstract

The two primary issues of historical data are cost and volume. Historical data does not fit well with high-performance processing because historical data slows down the system. The best fit fort historical data is the data warehouse. There are lots of reasons to keep historical data. The most basic reason is that consumers are creatures of habit. By knowing the historical patterns of behavior of a customer, the organization can predict future patterns of behavior.

Keywords

historical data

patterns of behavior

creatures of habit

transaction processing

data warehouse

Keeping data over time has always been problematic. For a variety of reasons keeping a historical record of data has had its problems. Figure 3.3.1 ...