Printer Friendly


You may think you're using your data warehouse to its fullest extent. You've amassed a great deal of information about either your customers or your inventory and you're able to look up information at will. However, without the use of the proper Business Intelligence tools, your data warehouse is nothing better than a hunk of coal. Using BI tools, like data mining, you can turn that coal into the diamond that can propel your business to the top.

"There is no question that data mining can help a company do more than they ever realized with their data warehouse," says Michael Pelletier, a consultant with Pinnacle Decision Systems, a professional services and software development company in Middletown, Connecticut. "Some call data mining 'data discovery' or 'data knowledge.' However you slice it, data mining allows you to take your data warehouse, dissect it piece-by-piece, and then summarize it into useful information. For example, you can use it to find out how many and what type of customers have been buying a certain product and then use it to help predict future sales for the product."

Data mining isn't just a fancy term for Online Analytical Processing (OLAP). OLAP tells you what happened in the past. Data mining helps users predict the future through the discovery of trends and patterns in the data. This insight can then help a business repeat good trends and avoid bad trends. Most business users will be utilizing a combination of both.It's important to keep in mind the difference so that the proper tool is used to answer the question at hand.

"One of the nice things about today's data mining tools is you don't need a computer science degree to use them effectively," says Pelletier. "Some of the latest data mining tools are so simple that most business users can use them effectively. Even though the complex algorithms are hidden, you still need to have some understanding of why and what you're doing with the data mining tool, but overall, the programs are much more user friendly than they used to be."

Data mining tools are especially important for marketing departments because marketing efforts succeed when they single out customer segments with high profit potential. Before, it was difficult to identify these groups because it required sifting through large amounts of data manually. With data mining, it's easy to sort through the data and discover customer behavior and patterns quickly.

"One of our customers in New York was an online company which was examining the effectiveness of their Internet advertising campaign. They wanted to find out why people were coming to their site and who was actually buying products," says Tom Todaro, a consultant in Pinnacle's New York office. "We were able to use a data mining tool from Cognos that allowed them to sort through their data warehouse to profile their customers in such a way that they were able to find out which referring web sites gave them sales and which ones didn't. That way they were able to better target their future marketing efforts."

Data mining can also directly affect a business' bottom line by improving quality. The analytical features of data mining allow business leaders to identify best practices that are hidden in their data, institute fraud control and support initiatives that center around operation improvement. Data mining tools are new weapons that weed out inefficiencies and help a company outpace the competition.

There are a number of data mining tools out there. Some of the more popular tools are Cognos ScenarioTM, Angoss Knowledge Suite, Business Objects Business Miner, SAS Enterprise Miner and SPSS Clementine. Microsoft(r) SQL Server 2000, which is due out in a few months will have data mining features. There are also related tools that do predictive modeling or present the results of the data mining graphically. But which program is best? There really is no answer that that.

"The best program for you may not be the best program for another company," says Todaro. "It really depends upon the level of management and sophistication you need. If you're a visual person who needs to see it in a graph, you want something like Cognos VisualizerTM. If you're a person with a bunch of thoughts, but not sure where they lead, you may want to use Cognos ForethoughtTM."

To best use data mining tools, Todaro says there are a number of steps. "First clean up your data warehouse. The old saying of 'garbage in, garbage out' still applies," he says. "Once you've cleaned up your databases, integrate all of them into one database. This is usually done using some sort of Extraction, Transformation and Loading (ETL) tool. Then, figure out which type of data you are going to use. Before going forward, it helps to trim off the fat as much as possible."

Once you've gotten your core listing of data cleaned and ready, Todaro says you need to transform it into the appropriate form for use by the mining tool. Once that data is transformed, you're ready to start mining.

"This is when you use a pattern finder to figure out what types of patterns exist in your data," Todaro says. "Once you find some patterns, your last step is to figure out which of those patterns is useful and interesting and which ones are aren't. It may take a while to find out which trends are useful to your business."

The use of data mining is still up and coming. But it's becoming clear that without data mining tools, businesses aren't getting all they can out of their data. The growth of e-commerce has helped data mining catch on because web-based companies are amassing huge amounts of data on their customers. Typically, when someone visits an e-commerce site, they leave an electronic trail behind which tells what they looked at and what they did. All of the page hits lead to massive amounts of data that would be impossible to analyze with traditional tools. However, using data mining tools, this data can be turned into gold.

"I always recommend that businesses consult with a professional IT company when they are considering data mining," says Pelletier. "It's best to contact a professional very early on in the data gathering process so that you can figure out the best way to store your data. Remember, storage is cheap and it's very hard to determine from the outset what information is important. What you think may be worthless, may turn into the diamond you need. I always say work with a professional to store as much information as you can and only get rid of it when you know it's useless."

Pelletier also says when you're thinking of starting data mining it's a good idea to start small and then work up. "Don't try to create the world in one shot. It isn't a bad idea to start small and use data mining to focus on one business issue, like retaining customers," he says. "You will have more chance of success by concentrating on one issue to meet your long term goal."

Lastly, Todaro says you must remember to keep the key decision makers involved during the entire data collection and mining process. This is because they will be the ones that are using the data and they need to know how to use it. "Think of decision makers as the miners with the helmets and lights," he says. "You can send them into a cave but if you don't tell them what to dig for and how to do it, they won't find anything."

Pinnacle Decision Systems is a privately held professional services and software development company that provides complete, creative IT solutions for information management needs. The consulting division uses the latest technologies to design and develop Internet-based business intelligence systems incorporating client/server and data warehousing applications. It creates custom software for Fortune 1000 companies,'s and government agencies with an emphasis on decision support systems. The company's software development division designs, develops and markets software products including HQ Intranet - The Company Communicator, Golf Insight, an on-line golf improvement system, and PinnPoint Plus - a Utility Tool for 4GLs. Pinnacle Decision Systems is headquartered in Middletown, Connecticut with additional offices in New York and Boston. Its Web site is located at

For more information on data mining, call Pinnacle Decision Systems at (860)632-7766.
COPYRIGHT 2000 Worldwide Videotex
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2000, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Technology Information
Publication:Online Product News
Geographic Code:1USA
Date:Nov 1, 2000

Related Articles
PlanView Web Software Version 7. (Products).
Using BI tools to turn information into action. (Special Report: Business Intelligence).
SAS sets new standards on 64-bit BI support.
CFOs positioned to drive BI integration: two major CFO challenges--improving performance management and improving access to information--can be...
MySQL V5--ready for prime time business intelligence.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters