ELECTRONIC GOVERNMENT AND THE POSSIBILITY OF DATA MINING
By Ed Colet
An article in the New York Times, 9/30/1999, page G1, "For the People, by the Computer", describes recent efforts by various government agencies to move on-line to deliver services to the public. This concept of electronic government relies on the Internet to provide people with access to municipal services and information. These services include a wide range of activities such as the ability for citizens to register complaints, pay parking tickets, pay taxes, apply for permits and track the approval process of these applications, review budgets and even send feedback and suggested changes to budgets. Since much of this activity is only at the preliminary stages of being implemented, there are some logistical implementation issues, and social equity concerns that need to be addressed. But beyond these concerns are a myriad of benefits. In this column, I project how data mining could possibly be integrated with an electronic government.
A move to electronic government represents a paradigm shift accompanied by issues that are similar to those that corporate and industrial domains had to address when they first moved on-line. For example, the banking industry went through a paradigm shift when it moved on-line away from people conducting transactions in person via a teller window - would this adversely affect those customers without on-line access? In the e-commerce domain, there are ongoing concerns about privacy and security. As government moves on-line, these same issues have come up as well. But just as data mining has now become an integrated aspect of both banking and e-commerce, data mining can also become an integrated aspect of electronic government.
An electronic government requires on-line systems to store and record transactions between the public and the government and it's possible that these databases can then be mined to reveal subtle patterns. Through data mining, the public's concerns can be analyzed and better government services can be provided. For example, frequent complaints about street cleanliness (the lack thereof) may warrant more frequent street cleaning services for a certain section of the city. Data mining can also reveal if specific sub-segments of the population have distinctly different concerns and expectations, resulting in uniquely different programs and services being made available - more youth recreation programs here, more adult activity programs elsewhere. In the corporate domain, data mining is used to personalize and target products to individuals. But should government be in the business of providing different and "uniquely personalized" services only to selected individuals based on data mining? Would this be contrary to spirit of a democratic government? (All interesting questions, but beyond the scope of this column).
The above examples assume that the data mining is conducted by end-user(s) that are government officials seeking to improve services. But it is also possible to imagine that the end user of a data mining application is the public citizen. It would then be possible to analyze in detail the voting patterns of an elected official(s) on only those issues that are important to the concerned citizen.
Unlike the corporate domain in which highly trained analysts perform data mining, in the concept of an electronic government, the end-users (either government officials, or public citizens) are considered to be lay-persons. As such, ease of use, and intuitive data mining interfaces utilizing a web browser are necessary. Building this is entirely possible today (some of our current work at Virtual Gold is exactly along these lines).
But before data mining can become an integral part of electronic government, a very large-scale systems integration problem has to be solved. The scope is much larger than in any corporate setting - local, city, state, and national government agencies present a mixed bag of hardware, software and related equipment. Developing a data mining application that can tap into databases across these levels is unrealistically optimistic at this point - although the concept is appealing. In the end, one benefit would be a more interactive relationship between government and an actively involved citizenry.
Ed Colet is the Acting Director of Research at Virtual Gold Inc., responsible for developing analytical methods for data mining and for investigating human factors and usability issues of business intelligence systems. At present, he is in the final stage of completing a doctoral dissertation in the Cognition and Perception program at New York University's Department of Psychology. Ed has also worked for IBM Research at the T.J. Watson Research Center. At IBM, Ed was a member of the group that developed Advanced Scout, the data mining application for NBA teams. His research interests focus on statistical methods and human factors.
For more information, see http://www.virtualgold.com.