Data mining
By Maggie Le
Data Mining: What is Data Mining?
Data analysis process from a variety of sources and abridging into helpful information that
can be used to, e.g. raising income and reducing cost in business
Software for data mining are analytical tools in data analysis since they can help to analyze
data in many ways, organizing and regrouping data, and put in brief identified relationships
between data. On a technical basis, data mining is itself the process of searching for
connections within big numbers of fields in vast relational bases.
For example, a regular grocery group used Oracle software’s data mining capacity in the
analysis of local purchasing patterns and from that they found out that: people buy diapers
and beer on Thursdays and Saturdays, buy little on Thursdays and do their grocery
shopping on Saturdays, and conclusion is made by the retailer: people buy beer for their
weekends. This kind of info. can be used to raise income such as in the moving of beers
next to diapers on display, and increasing the price of beers and diapers on Thursdays to
full price
Uses/applications
The use of data mining at present:
- Companies (in retailing, finance, communication and marketing bodies) that put their
customers as first in their business employ data mining to help them in figuring out links
concerning intrinsic factors e.g. price, placement of product or staff aptitude as well as
those concerning the extrinsic factors e.g. indicators of the economy, corporate rivalry
and customer stats. Not only that, data mining enables businesses to resolve sales
outcomes, customer expectations and collective gains as well as enabling them to dig
into laconic information to see conventional data in detail.
- Thanks to data mining, point-of-sale records from what customers purchase are used
by retailers in the sending of objective promotions that are based on the purchase history
of a customer. Developments of products and promotions are made by retailers to entice
particular subdivisions of customers thanks to the mining of poll data based on comment
or assurance cards.
- Recommendations of rentals to customers are established by Blockbuster
Entertainment from their mining of database of video rental history.
- Walmart is inventing large data mining to change its drug dealer connections. Walmart
gets hold of point-of-sale conventions from more than 2,900 shops in six nations and
send the data non-stop to its large 7.5 terabyte, Teradata data storehouse. Walmart let
over 3,500 drug dealers get access to their product data and execute analysis of data.
These drug-dealers utilize the data to recognize patterns in customers’ purchases at the
shop front zone. This information is employed in the management of local shop stock and
look for latest business chances. In the year1995, one million mixed data problems were
treated by Walmart computers.
Ethical/social issues
Problems in the supplying of raw data for processing within internal storage
done through databases : databases are potentially dynamic, non-organized
and massive.
The data mining issue at first maybe: databases which are usually designed for
uses distinct from that of data mining and at times the characteristics or
features that would make the research work easier are not existing nor possible
for demand from reality. Infinite data are complicated due to the fact that
ifcertain features important to the understanding the area of application are not
found in the data it can be hard to find out convincing facts on a particular area,
e.g. A patient cannot be diagnosed for malaria from a database for patients
unless the database has the count of the patient’s red blood cell.
Limit in information available
Bibliography
http://www.pcc.qub.ac.uk/tec/courses/datamining/stu_notes/d
m_book_2.html#HEADING18
http://www.anderson.ucla.edu/faculty/jason.frand/teacher/tech
nologies/palace/datamining.htm