Data Warehousing

Getting the most from your data

Areas such as Data Mining, Data Warehousing and Database Marketing are becoming increasingly important for OR. Here is a quick guide to the state of the art software in these areas

by Nigel Cummings

Knowledge management is firmly embedded in management thinking for the nineteen nineties, and the use of intellectual capital has never been a hotter topic. Yet the subject is not a new one; the whole concept of business intelligence has been around for over twenty years. According to a recent report in Strategy, the official journal of Business Intelligence, 1976 was the time when Ben Heineman, chief executive of Northwest Industries, decided to use a computer terminal and monitor to plan the growth of his operating units. Heinemann and others of his day first saw the potential for the use of information technology by senior management.

But the seventies were not a good time, technologically speaking. Desktop computers were few and far between, computing power was low, and equipment costs were high. The beginning of the nineteen eighties provided a much more equable climate for the utilisation of information technology. Cheap CP/M based systems with decent specifications proliferated from companies such as ACT and IBM, and other major companies were mindful of the evolution of technology to new levels of performance.

The apex of this evolution seemed to occur in the early eighties when the DOS based Personal Computer (PC) was born. Initially parented by IBM, the PC bred quickly and became established across a wide licence base as the number one business machine. Looking at the early personal computers from a nineteen nineties perspective, it is hard to see how we managed to gain any productivity benefits from them - they were in comparison with today's computers feeble minded and lacking the graphics capabilities which we now take for granted.

Around the time of the first PC, EIS (Electronic Information System) appeared. EIS was the forerunner of today's Business Intelligence systems, and it may be considered that we owe EIS a debt of gratitude for providing the route to the graphics oriented business systems we know today. The term Business Intelligence (BI) was first coined by the US based Gartner Group as a concept and core topic of research. From a strategic planning perspective, by the year 2000 Gartner predict that 'Information Democracy' will emerge in forward thinking organisations and BI data and applications will be broadly available to employees, consultants, customers, suppliers and the public.

BI enablement is being achieved as more companies take on board query and reporting tools, EISs and decision support applications. Information Democracy (ID) bears similarity to political democracy which is intended to ensure equality for all citizens. ID on the other hand will promote the common good of an organisation, though not necessarily the individual. Users will be guaranteed access to information, but they will not all have the same level of access nor access to the same information.

Information quality will have to be considered, as data warehousing activities have now gained the interest of information system organisations. The key factors in BI are, according to the Gartner Group: timeliness, relevance, consistency, and completeness. The group feels that the outcome of an information quality architecture should serve as input into any data warehousing effort that an organisation is undertaking. As a precursor to BI applications development, IS organisations and end users should be encouraged to maintain an information quality architecture as part of an overall BI architecture, thus improving decision-making quality by more closely mapping business needs to available information

First published to members of the Operational Research Society in Inside OR August 1998