Page 160 - DCAP606_BUSINESS_INTELLIGENCE

Unit 11: Data Mining

of product selection and placement conclusions, coupon offers etc. Companies such as phone Notes
service providers and music clubs can use data mining to create a “churn analysis,” to assess
which customers are expected to stay as subscribers and which ones are likely to switch to a
competitor.
In the public sector, data mining applications were initially used as a means to detect fraud and
waste, but now they are used for purposes such as measuring and improving program
performance.

Self Assessment

Fill in the blanks:
4. Information can be converted into .................................
5. ............................ refers to removal noise and inconsistent data.
6. .......................... method uses some variable to predict unknown or future values of other
variables.

11.4 Data Mining Issues

Privacy

One of the key matters raised by data mining technology is not an enterprise or technological
one, but a social one. It is the issue of individual privacy. Data mining makes it possible to
investigate routine enterprise transactions and glean a significant amount of information about
persons buying habits and preferences.

Data Integrity

Clearly, data analysis can only be as good as the data that is being analysed. A key implementation
dispute is integrating inconsistent or redundant data from distinct sources.

Example: A bank may sustain credit cards accounts on several distinct databases. The
addresses (or even the titles) of a single cardholder may be different in each. Software should
convert data from one system to another and choose the recently entered address.

Confusion

Another issue of concern is to decide is if it is better to set up a relational database structure or
a multidimensional one. In a relational structure, data is retained in tables, allowing ad hoc
queries. In a multidimensional structure, on the other hand, groups of cubes are arranged in
arrays, with subgroups created according to category. While multidimensional organisations
facilitate multidimensional data mining, relational structures perform better in client/server
environments.

Cost

Finally, there is the issue of cost. While system hardware costs have fallen spectacularly inside
the past five years, data mining and data warehousing are inclined to be self-reinforcing. The
more mighty the data mining queries, the larger the utility of the data being gleaned from the

LOVELY PROFESSIONAL UNIVERSITY 155