Data Mining

What Does the Term "Data Mining" Mean?

Data mining is the process of analyzing hidden patterns in data from various perspectives to turn that data into useful and often actionable information. The data is collected and stored in common locations such as data warehouses, and data mining algorithms look for patterns that can be used by the enterprise to make better decisions, such as those that reduce costs, increase revenues, or improve customer service.

Data mining is also known as data discovery or knowledge discovery. It is crucial in business intelligence for data-driven decision making.

The main steps of the data mining process are:

• Extracting, transforming, and loading data into a data warehouse
• Storing and managing data in multidimensional databases
• Providing data access to business analysts via application software
• Presenting the analyzed data in an understandable form, such as charts.

The first step in data mining is collecting relevant data that is important for the business. Company data can be transactional, non-operational, or metadata. Transactional data concerns daily operations such as sales, inventory, and costs. Non-operational data is usually predictive, and metadata is related to logical database design. Patterns and relationships between data elements often yield the necessary information for improving business processes. Consumer-focused organizations use data mining techniques to gain a clear understanding of products sold, pricing, competition, and buyer demographics.

For example, retail giant Walmart transmits all necessary information to a data warehouse containing terabytes of data. This data is easily accessible to suppliers, enabling them to identify purchasing habits. Using data mining techniques, data on buying habits, most frequently visited stores, most sought-after products, and other insights can be obtained.

The second stage of data mining is selecting an appropriate algorithm—a mechanism that creates a data mining model. The general function of the algorithm is to identify trends in the dataset and use the results to determine parameters. The most popular algorithms used for data mining are classification and regression algorithms, which are used to identify relationships between data elements. Major database vendors such as Oracle and SQL include data mining algorithms like clustering and regression trees to meet the demand for data mining.