What is data mining? What's the use?

Defined

Data Mining is the nontrivial process of acquiring effective, novel, potentially useful, and ultimately understandable patterns from large amounts of data stored in databases, data warehouses, or other repositories.

What is the use of.

Data mining, simply said that there is historical data, a lot of data, such as the bean has accumulated a lot of user data, if there is a user, like to listen to songs, like technology, like what group, but participate in, speak, label, etc., these data can enter a model of data mining, select algorithms, Analysis, So a lot of customer behavior found that some of the behavior is to use the eyes to see the data can not be seen, and these rules or discoveries are called knowledge.

For example, in the watercress, about the recommendation of the book, keywords or tag similarity calculation, such as someone bought data mining, but also to buy tea books, these 2 keywords are different from the classification point of view is connected, but from the point of view of data mining, the user's habits so.

In the site of intelligence and personalization also need data mining technology support, such as Taobao, according to the user's search habits, the introduction of users like products.

Mining objects

In principle, data mining can be carried out on any type of data, which can be commercial data, which can be data from the social sciences, natural science processes, or satellite observations. Data form and structure are also different, can be hierarchical, network, relational database, can be object-oriented and object-relational advanced database system, can be oriented to special application of the database, such as spatial database, time series database, text database and multimedia database, can also be web information. Of course, the difficulty of data mining and the technology used also vary depending on the data storage system.

Mining methods

Association rule method, decision tree algorithm, neural network method, rough set algorithm, genetic algorithm.

Mining results