Definition of

Data mining

The concept of data mining , coming from the English language, is usually mentioned in our language as data mining . The notion is linked to the procedure carried out to detect patterns in a large amount of data.

Data miningThe purpose of data mining is to extract certain information from a mass of data to create a structure that can be understood and used. For this, it uses database systems, statistical techniques and other resources.

Data mining analyzes and processes data in search of some pattern or model . Once a certain structure is discovered, it seeks to make it visible so that it is possible to work with it.

In this way, through a semi-automatic or automatic analysis of the data , data mining manages to discover patterns that, until now, were not known. From then on, additional tasks or activities arise that, although they do not belong to the specific scope of data mining, are part of its universe.

It can be said that the data mining process begins with the selection of the mass of data. Then we proceed to analyze their properties to transform them and extract information that can be interpreted and evaluated.

More precisely, we can state that data mining is made up of three clearly defined stages or phases:

-Determining the objectives. That is, it is about establishing what purposes are pursued with that process. It is the person who orders them who decides them and then makes them known, logically, to the data mining professional.

-Data preprocessing. It consists of the selection, cleaning, enrichment, reduction and even change in the databases that are key in the process.

-The choice of the model. This stage, in turn, we can establish that it is divided into several parts. Thus, first of all, the statistical analysis of the data is carried out. And then, secondly, the graphic visualization of those is developed.

-The analysis of the results obtained, with which it will be possible to know what point has been reached and also whether the purposes intended by whoever commissioned the data mining process have been achieved.

Data mining can be used, for example, to detect possible terrorists . Through the analysis of millions of telephone calls, emails and communications of different types, it is possible to discover some pattern that allows us to identify people who plan to commit an attack.

A company can also use data mining to search for certain variables among the data it has on its customers and thus offer a certain product only to those who meet certain requirements.

Precisely in that sense, supermarkets and hypermarkets can find a great ally in data mining. And, for example, if thanks to this they know the habits of their clients regarding the weekend, they can be clear about which products are most consumed during that period and thus make them more attractive or even make them more within reach. . In this way, they will satisfy customers and improve their sales.