Data mining is the process of extracting information from large sources of data, such as a corporate data warehouse, and extrapolating relationships and trends within that data. It is not possible to use standard query tools, such as SQL, to perform these operations. There are three main categories of data mining tools: query-and-reporting tools, intelligent agents, and multidimensional analysis tools.
Query-and-reporting tools offer functionality similar to query and report generators for standard databases. These tools are easy to use, but their scope is limited to that of a relational database, and they do not take full advantage of the potential of a data warehouse.
The term 'intelligent agents' encompasses a variety of artificial intelligence tools which have recently emerged into the field of data manipulation. Two of these tools are neural networks and fuzzy logic. An intelligent agent can sift through the contents of a database, finding unsuspected trends and relationships between data.
Multidimensional analysis tools allow a user to interpret multidimensional data (i.e., a hypercube data set) from different perspectives. For example, if a set of data includes products sold in various regions over time, multidimensional analysis allows you to view the data in different ways. For instance, you could display all sales in all regions for a given time, or all sales over time in a given region.
|
The use of intelligent agents in online analytical processing (OLAP) is sometimes known as "information discovery". These tools can process extremely large amounts of data and interpret it in a manner that approximates human decision-making, inferring relationships between data that might never be found by traditional analytical processes. |
Data mining tools are used to retrieve information from data warehouses. There are three basic types of data mining tools: