What are the features of data mining?

There are various features of data mining that are as follows −

Data types − Most data mining systems that are accessible in the industry handle formatted, record-based, relational-like data with statistical, categorical, and symbolic attributes. The data can be in the form of ASCII text, relational database data, or data warehouse data. It is essential to test what exact format(s) each system it is treating can handle.

Some types of data or applications can require specialized algorithms to search for patterns, and so their requirements cannot be managed by off-the-shelf, generic data mining systems. Rather than, specialized data mining systems can be used, which mine either text reports, geospatial data, multimedia data, stream data, time sequence data, biological data, or Web data, or are dedicated to specific applications (including finance, the retail industry, or telecommunications).

System issues − A given data mining system can run on only one operating framework or several. The famous operating systems that host data mining software are UNIX/Linux and Microsoft Windows. There are also data mining systems that run on Macintosh, OS/2, etc. Large market-oriented data

Large market-oriented data mining systems often adopt a client/server architecture, where the client can be a personal computer, and the server can be a collection of powerful parallel computers. A current trend has data mining systems supporting Web-based interfaces and enabling XML data as input and/or output.

Data sources − This defines the specific data formats on which the data mining system will operate. Some systems run only on ASCII text files, whereas some work on relational data, or data warehouse data, accessing several relational data sources.

A data mining system must provide ODBC connections or OLE DB for ODBC connections. These provide open database connections, especially, the ability to access any relational data (involving those in IBM/DB2, Microsoft SQL Server, Microsoft Access, Oracle, Sybase, etc.), and formatted ASCII text data.

Data mining functions and methodologies − Data mining functions form the heart of a data mining system. Some data mining systems support only one data mining function, such as classification. Others can help multiple data mining functions, including concept description, discovery-driven OLAP analysis, association mining, linkage analysis, statistical analysis, classification, prediction, clustering, outlier analysis, similarity search, sequential pattern analysis, and visual data mining.

For a given data mining function (including classification), some systems can provide only one method, whereas others can provide a wide variety of methods (including decision tree analysis, Bayesian networks, neural networks, support vector machines, rule-based classification, k-nearest-neighbor methods, genetic algorithms, and case-based reasoning).

Data mining systems that provide multiple data mining functions and multiple methods per function support the user with higher flexibility and analysis power. Some problems can require users to try a few different mining functions or incorporate several together, and different methods can be more efficient than others for different kinds of data.