Index Last Week Contact DQ NEWS Latest Issue
MIT Conference: Don't 'TQM' Data
and
Information Quality!
Among the most heated discussions at the MIT Information Quality Conference (October 24-26) were those about the future of information and data quality. Several conference participants who are long-time Quality practitioners voiced their concerns.
Chief among these concerns was the feeling that DQ and IQ are about to go down the same path as "Total Quality Management," i.e., DQ and IQ have the potential of becoming corporate "buzzwords" like "TQM." The concerns voiced at MIT were: where would that put us, and what would that get us? As most Quality practitioners are aware, while executives, consultants, seminar givers, and Quality Gurus were estatic about TQM, many at the bottom of the corporate and organizational ladder came to despise TQM. Why? All too often work piled-up on mid-level and low-level employees' desks while they atttended TQM seminars. Six months after the TQM seminars, these employees felt the seminars were a total waste of time - the seminars had little or no positive effect on the organization. Several participants likened TQM seminars to religious "revival meetings." Several other conference participants stated that DQ and IQ "revival meetings" are already being held.
A similar concern voiced at MIT involved an organization's ability to improve its information and data quality. One participant pointed out that, all too often, database functions are performed by persons who are poorly paid, poorly trained, and poorly motivated. Can organizations significantly improve information and data quality given these constraints? Can gurus, seminars, consultants, and conferences really "help" such organizations? A participant with a background in Industrial Quality pointed out that it was only through a significant positive change in the manufacturing sector's "corporate culture" that a significant improvement in product quality occurred. He maintained that Quality consultants had spent years doing "process capability" studies on obsolete manufacturing systems, before a crisis in American manufacturing changed the focus from product inspection/repair to process improvement. For information about the 1998 MIT Information Quality Conference please go to: http://web.mit.edu/tdqm .
DQ Journal and News to be Redesigned, Expanded
The editor of journal DATA QUALITY mailed the printed version of the September issue to the Editorial and Editorial Advisory Board for their comments. The editor will begin putting the on-line version of the journal on-line about December 1st. The publisher of DATA QUALITY has hired a graphic designer to improve the journal's design and layout. Starting November 9th, the weekly on-line newsletter will be expanded to cover more topics. To send us your comments, click here: COMMENTS
Project Seeks World-Wide Internet Data Standards
According to an article in the November 3rd issue of The New York Times, a world-wide effort is underway to get government and commercial information providers to agree to a global standard for providing access to information. The project, named the Global Information Locator Service (GILS), was developed by the U.S. Geological Survey in Reston, Virginia. The USGS is part of a cooperative effort that includes hundreds of experts in over a dozen industrialized nations.
With GILS, instead of visiting many Web sites to search a particular topic a user can search many sites from one location. For example, at the Fedworld Web site (www.fedworld.gov/gils) a user can search across several Federal agency collections at the same time. GILS is like a card catalog that both describes agency resources and provides assistance in obtaining information. GILS works with text search engines, but can also access photographs, films, and complex information like chemical formulas. [Editor's Note: GILS adopts the international standard for information search - ISO 10163, known in the U.S. as ANSI Z39.50. The U.S. Library of Congress is the official maintenance agency for Z39.50. See the Library's Web site: (www.loc.gov/z3950/) and the USGS Web site: (www.usgs.gov/gils/locator.html)].
Another important difference between GILS and Web indexers ("crawlers") is that Web indexers can index static web pages, but cannot easily index information that lies within a data base. GILS is described as a way of applying succinct and user-friendly descriptions to information resources, particularly at an aggregated level. Many of these sources are big, rich, and complex; and are not just a collection of static Web pages. Examples of sites that use the GILS system can be found at (www.usgs.gov/gils/showcase/).
A Federal mandate requires all Cabinet-level agencies to be GILS compliant. Some big Federal agencies, like the Defense Department, are more GILS compliant than others. Among the 5,000 sites that are using the GILS search standard are those at AT&T, Ameritech, and Hyundai. The article was written by Times information technology writer Sreenath Sreenivasan and appears on page D5.
Climate Models - Was Data Quality Disregarded?
According to a lengthy report in the November 4th "Science Times" section of The New York Times, climate scientists have developed increasingly sophisticated ocean-atmosphere climate models during the past decade. Over 20 ocean-atmosphere climate models now exist. Simulations using these models are the basis for the predictions that - unless there are reductions in greenhouse gas emissions - the earth's average surface temperature will rise substantially and disrupt the world's climate in the coming decades.
According to the Times, all the ocean-atmosphere models thus far developed predict a global rise in temperature but the predictions vary regarding the amount of warming. The models are not good predictors of local warming trends. The data used in the models consist of values assigned to "cubes" of a three dimensional grid typically rising 10 or 12 miles above the earth. The typical grid spacing is less than half a mile vertically and 150 miles horizontally. Data may consist of assigned or observed values, or a combination of both. One stated problem is the sparseness of observed data for a number of variables, such as water vapor. Another problem is that there may be "missing" variables that significantly affect global and local climate.
Data and information quality - and their effects on climate models - are not specifically mentioned in the article, which was written by William Stevens and appears on page C1.