Database Administration

Small businesses also have to make sure that end-users from every department are actively involved in the design of the data warehouse. Without that type of feedback, the database may turn out to be useless because it does not store the right type of information, or it stores it in the wrong way. Those end-users involved in the design can learn how to use the system first and then serve as trainers in their department. Once the main warehouse is built, it then becomes easy for each department to build its own data marts by pulling out the fields from the main database that it needs.

The initial costs of building a data warehouse are high—software, hardware, and consulting fees add up quickly. However, most businesses, from supermarkets to banks, are finding that having a data warehouse is a competitive necessity. One example of how data warehouses are being used is a practice called "data mining." Data mining is the technique of creating statistical and predictive models of the real world based on patterns that are discovered as a result of complex queries performed on the huge amounts of data stored in a warehouse. When data mining is done right, it can produce amazing results, spotting trends before they happen or identifying new sales prospects, for instance. When done incorrectly, however, data mining can produce false correlations and misleading results. Companies should not rely too heavily on data mining and should ensure that they hire professionals who fully understand statistical analysis to perform the task.

DATABASE MANAGEMENT AND THE INTERNET OR INTRANETS

Data warehouses started out as internal projects, but now they are being seen as the next logical step on the Internet. Companies that need to gather data from customers and pass information down the line to business customers are finding it beneficial to make their data warehouse available over the World Wide Web. This means that either HTML or Java-based client servers need to be created to allow Web users to search the database. If the company desires to gather information on customers, it might make the warehouse available to the public over the Web. If the main purpose is to pass information on to business customers, then the company will make the warehouse available as part of its corporate intranet, which is available only to selected individuals.

At first, only basic queries could be run easily over the Web, but observers note that the situation has changed rapidly. Each of the major database management systems companies has scrambled to enable their databases to work closely with Web servers. Using new technology known as online analytical processing (OLAP), high-level, intricate queries of data warehouses will be possible. At the same time, consumers looking to data mine corporate information should be able to run simple queries.

There are risks associated with making such huge amounts of data available over the Web. Security is the paramount issue, since opening data warehouses to users around the world means that internal systems are exposed to outside interference or hacking. A second issue is the drain on system resources that unlimited access to the data warehouses would cause. Popular databases visited and searched by large numbers of users would need extremely powerful servers to keep up with demand. The servers would have to ensure that the employees and clients of a company would not be hindered by the excess traffic on the system. Finally, there is the issue of cost. As with any new technology, opening databases to the Web costs money. In addition to development costs associated with creating the search engines and OLAP tools, businesses will also have to weigh the cost of the powerful servers needed to meet the increased demand for information.

 PREV  1 | 2 | 3