| Index | Last Week | Contact DQ News | Latest Issue |
| Featured Articles | Data & Information | Science & Technology | Education |
| Economics | Business | Law | Medicine |
1999 Baldrige Award Criteria - Information Intensive
The U.S. Commerce Department recently published the criteria for the 1999 Baldrige National Quality Award. The Baldrige Award is a public-private partnership, which builds active partnerships in the private sector, and between the private sector and all levels of government. Among the organizations that play a role in the success of such partnerships are the Malcolm Baldrige National Quality Award Foundation, National Institute of Standards and Technology (NIST), American Society for Quality (ASQ), Baldrige Award Board of Overseers, Baldrige Award Board of Examiners, and Baldrige Award Recipients.
According to a brochure furnished by the U.S. Commerce Department, The Malcolm Baldrige Criteria for Performance Excellence are the basis for self-assessments, for making Awards, and for giving feedback to applicants. The core values and concepts of the Baldrige Award are embodied in seven categories: leadership, strategic planning, customer and market focus, information and analysis, human resource focus, process management, and business results.
The information and analysis category examines an organization's performance management system and how an organization analyzes performance data and information. (The term "information and analysis" refers to key metrics used by an organization to analyze and measure performance.) The information and analysis category consists of two components: measurement of organizational performance and analysis of organizational performance.
The Baldrige Award is judged by a panel of examiners, who assign scores to each of the seven categories based on the competing organizations' responses. The maximum score is 1000 points. The scoring is weighted to favor "business results" (450 points). The information and analysis category is worth 85 points. There is no mention of data and information quality in the descriptions of the Award criteria - data and information are expected to be "reliable." Although the Commerce Department's brochure explains the Award process in detail, it is uninformative about data and information quality vis a vis the examiners who assign scores to organizations that compete for the Award. Information about the Malcolm Baldrige National Quality Award may be obtained from NIST's Web site: http://www.quality.nist.gov.
Knowledge Discovery + Data Mining = Data Quality Problems
An article in the January-February 1999 issue of American Scientist discussed how "knowledge discovery " and "data mining" are being increasingly used by scientists to extract useful data and information from large scientific databases. One of the most important parts of a scientist's work is the discovery of patterns in data. Yet the databases of modern science are frequently so immense that they preclude direct human analysis. As data collection methodology has become automated, scientists have begun to search for ways to automate data analysis as well.
The volume of scientific data isn't the only problem, according to the article's authors. To make sense of data collected by satellite (for example) analysts will have to compare it with similar data collected from other sources. A third problem involves data interpretation. A satellite photo of vegetation may have to be calibrated to a known standard for a certain type of vegetation.
According to the article, data mining is just one part of the process of knowledge discovery in databases (often abbreviated KDD). The article's authors interpret KDD as an iterative process with six stages: 1) develop an understanding of the proposed application; 2) create a target dataset; 3) remove or correct corrupted data; 4) apply data-reduction algorithms; 5) apply a data-mining algorithm; and 6) interpret the mined patterns. This process is not necessarily sequential.
Unfortunately the article's focus is upon de facto "inspect-repair" activities, whether such activities are stated or unstated. Data mining methods include decision trees, neural nets, database segmentation, market-basket analysis, and deviation detection. These methods rely upon high quality data. The assumption of the article's authors is that high quality data is available, or that data can be conditioned to an acceptable level of quality. In reality, this is often difficult or impossible. The article also raises concerns about privacy invasion on a massive scale. In the future, everyone who uses a computer network may be subject to intense surveillance based on usage patterns. The article was written by Carla E. Brodley, Terran Lane, and Timothy M. Stough of Purdue University, and appears on page 54.
Gates: Data - Ubiquitous and 'Invisible' in 2009
In a featured interview in the January 4th issue of Computerworld, Microsoft Chairman and CEO Bill Gates forecast a future in which PCs are ubiquitous, with data synchronized and available on all kinds of devices whenever users need it. The interview took place at the Comdex/Fall '98 trade show in Las Vegas in November 1998. The interviewer was Computerworld editor in chief Paul Gillin.
According to Mr. Gates, personal computers will be available everwhere, in many different configurations by 2009. Paper may become an antiquated storage medium. Data storage and migration won't be a problem. Users will be provided with the data they need, when they need them. Mr. Gates doesn't believe that data will be a problem. He believes that in the future "You won't even think about where information really is. You'll just know that when you go to your home machine or your work machine, the files are there. The machine will be replicating information, and when you update it, the information will go up into a cloud and come down on other machines. Logically, the information will be in a cloud, but data will come to your machine, and the actual applications will be there. Storage management will be invisible. The operating system will do authentication and speech recognition and bring the files down. But you won't know where the server is."
Paul Gillin was the featured speaker at the 1998 MIT Conference on Information Quality. The article appears on page 28.
Go to: Data Quality Home Page
Comments: dqemail@aol.com (1999-1-3)