Data mining process steps in when we are converting data, which is in large amounts and collected in a fast way, to a meaningful information through various analysis. When we examine the definitions of data mining, one common point in those definitions are a lot of data is being held in a database and secondly, meaningful information is obtained from the data.

Data mining is the process where, based on the various data held in the database, we reveal information which wasn’t discovered before, and using them in the process of making decisions and realizing an action plan.

image

We can give various examples for the purpose of implementing data mining on customers in the product and service sector. The important objectives in data mining are determining the most profitable market segments, choosing the most profitable customers, determining the acceptance level of product or service which is offered to customer in a new campaign. Some examples about the advantages that can be obtained from implementing data mining is given below.

Data mining can be used to select a target audience and to determine how it will be offered to the target audience while creating a campaign program for a product or service. An actual application sample made for these decisions is described below with the statistical methods employed.
It can reveal which characteristics in a product or service impacts the customer satisfaction in what extent, and which characteristics make the products preferable.
By calculating the credit risk of the customers, it can be forecast which customers will fail to make their repayments on time. By analyzing the characteristics of people who hinders their credit card payments, make delayed payments or no payments at all, potential people who will experience the same situation can be detected.
An establishment can make an analysis about their customers who left them for their competitors, and obtain the characteristics of the customers who preferred their competitors. Doing so, they can make forecasts about customers who are likely to leave them, and develop strategies in order not to lose them and reclaim the lost customers.
By determining the most profitable customers, most profitable customers among the potential customers can be confirmed. By determining the profitable customers, you can make special campaigns just for them. The most costly customers can be turned into less costly customers. For instance, by determining the customers that makes the most banking operations, they can be directed to internet banking instead of branch banking, which is more costly.

The indispensible priority of data mining is for data to be correct and qualified. Each data mining application should start with that. After making sure that data can be trusted, we create models fit for specific purposes. We can summarize them as follows:

Estimator models
Classification
Time Series
Regression
Descriptive models
Clustering
Summarization
Association Rule Mining