02-06-2012, 04:05 PM
A Study of Data Mining and Information Ethics in Information Systems Curricula
Data [email protected] (Size: 688.67 KB / Downloads: 42)
INTRODUCTION
Many companies today, like Wal-Mart, store
much of their business and customer data in
large databases called data warehouses.
Their customers are not told the extent of
the information that is accumulated on
them, how long it will be kept, nor the uses
to which the data will be put. (Hays, 2004)
This data is subsequently analyzed to produce
new information to help the companies
evaluate business processes and customer
behavior. The technique usually used to do
the analysis is data mining.
Privacy
Privacy is not easily defined perhaps because
the notion of privacy has evolved over
time, and now means different things in different
situations and in different cultures.
Most scholars define three types of privacy
(Tavani, 2004). Accessibility privacy is freedom
from intrusion. Historically, this is the
first notion of privacy to be codified into law.
The Fourth Amendment to the U.S. Constitution
protects citizens from unreasonable
searches and seizures.
Privacy Guidelines
Although, as noted above, there are few
laws in the United States governing the use
of personal data, many of the existing laws
and businesses have used the Code of Fair
Information practices of the Organization for
Economic Cooperation and Development to
guide them in setting informational privacy
policy. The code is based on eight principles:
Collection Limitation, Data Quality,
Purpose Specification, Use Limitation, Security
Safeguards, Openness, Individual Participation,
and Accountability.
RESEARCH METHODOLOGY
The research methodology of the study consists
of five iterative stages of analysis.
In stage 1 a sample of 29 data mining texts
were chosen and analyzed by the authors of
this study, in December 2004 – February
2005, as mostly representative of primary
data mining texts in ABET certified schools
of computer science and information systems.
The authors were already knowledgeable
instructors and researchers in data mining
and information ethics. The authors
identified the textbooks in a joint analysis
that included diverse editors and marketing
representatives of leading mining publishers
and in an independent survey and analysis
of scholarly and practitioner sources on the
World Wide Web. The textbooks were identified
from 2004 – 2000 publication dates of
the publishers.
IMPLICATIONS OF INITIAL STUDY
“We are not just heading towards a world of
Big Brother … but also toward a more mindless
process … - a world that is beginning to
resemble Kafka’s vision in The Trail.” (Solove,
2004)
The inadequacy of contemporary data mining
texts, for instructors attempting to effectively
introduce information ethics in computer
science and information systems curricula,
is an important implication of this current
study. Ethical principles for moral and
philosophical theory (Grodzinsky, 1999) continue
to be not included in depth in data
mining texts. Ethical and non-ethical practices
of citizen and consumer information
mining that are current and immediate in
government and industry are not often included
in outdated texts (Schrage, 2005).
Though instructors may enhance core foundational
texts with further practitioner and
scholarly resources, they are challenged in
having comprehensively convenient and integrated
texts (Richards, 2005).