Thursday, 21 November 2013

Ethical issues of Privacy Preservation in Data Mining vis-a-vis Quality of data

Abstract--Customer Relationship Management (CRM) is a business activity through which the policies and
processes, based on information technology can acquire and manage customer knowledge. CRM system in the
enterprises can collect, track and analyze information of every customer, further more understand customer needs
and meet them. Various applications of CRM like Profile analysis, Segmentation, Risk analysis, Attrition, Lifetime
Value of the customer etc. , need to look into the personal data for data prediction and finding trends.
Research in data mining has been done on discovering knowledge about the underlying data. Data mining tools are
used to maintain the data in data warehouse and predict the hidden patterns and present them in form of a model.
Strategic decisions about the customers can be taken based on the outcomes of these models. This gives rise various
ethical issues like invasion of privacy, deception, harm to the customers data, loss of customers identity etc. Privacypreserving
data mining (PPDM) deals with the trade-off between the effectiveness of the mining process and privacy
of the subjects, aiming at minimizing the privacy exposure with minimal effect on mining results.Privacy
preservation in data mining is a framework which allows computation of data over an entire data set without
compromising the privacy of the data of the customer. Most of the methods for privacy computation use some
transformation on the data in order to perform the privacy preservation. Though the issues of privacy preservation is
of data might be taken care but quality of data is also of concern.
This paper aims to review the existing literature regarding previous techniques for Privacy-preserving data mining
(PPDM) and then the researcher proposes to find the effectiveness of the quality of data under the paradigm of
privacy preservation during data mining and ethical issues which may arise due to it . Finally analyze the

observations and discuss the conclusions and future work.

No comments:

Post a Comment