Data mining is a term for the computer implementation of a timeless human activity: It is the process of using automated methods to uncover trends, patterns, and relationships from accumulated electronic traces of data. Data miningor knowledge discovery, as it is sometimes calledlets you exploit an enterprise data store by examining the data for patterns that suggest To mine data, you need access to data, so it’s no coincidence that data mining developed at the same time as data warehousing did. As computer power and database capability grew through the late 1900s, people began to see that data wasn’t simply a passive receptacle, useful only in performing billing or order-entry functions. People could also use data in a more proactive role to provide predictive value in guiding their businesses forward. This notion led to the development of a new breed of computer systems that went beyond running the business (as early computer applications did) to informing and analyzing the business. These new systems were sometimes called decision-support systems or executive information systems (EISs).
These systems were designed to harness growing computing power and improved GUIs to provide ad hoc analytical reports that could slice and dice data in novel ways and went well beyond earlier notions of static reporting. Slicing and dicing datadrilling down into detailed reports or zooming up to a 10,000-foot “big picture” viewrequired special ways of organizing data for decision making. This need gave rise to the data warehouse. The term data warehousing was virtually unknown in 1990. Ten years later, data warehousing has become a multibillion-dollar business of capturing and organizing data to provide a proactive analytical (versus operational) environment that uses data in defining and guiding business activity. As data warehousing matured, decision support and EIS gave way to the more general concepts of BI and data mining. BI involves organizing data along various potential dimensions of analysis so that you can cross-reference and display any view of datasay sales resultsfrom within any number of other potential dimensionssay region or product line. The ability to move up and down dimensions lets you drill down into detail or zoom up for a more general view.
The Business plan on Data Warehouse Information Database Analysis
... and summarised data.2. 8 BUSINESS USE OF A DATA WAREHOUSE No discussion of the data warehousing systems is complete without review ... data into computer memory which helps to significantly improve performance by minimizing physical disk I/O. In conclusion OLAP servers logically organize data ... finding patterns and regularities in sets of data. It is the computer, which is responsible for finding the ...
The ability to show variations in data along various dimensionsoften, many dimensions simultaneouslyprovides multidimensional reporting capability in realtime. This general approach to manipulating data became known as online analytical processing (OLAP)that is, processing data for analytical purposes instead of operational purposes. The term online refers to having the analytical data continuously available. OLAP takes advantage of a data warehouse by making data continuously available in a form that supports analytical decision-support tasks. The distinguishing characteristic of OLAP is the preprocessing, indexing, and storage of data in various dimensional representations to quickly deliver the various dimensional views BI requires. However, BI OLAP tools might not find all the patterns and dependencies that exist in data. OLAP cubes are appropriate for a limited amount of data exploration, involving major variations according to critical and known business dimensions. But when the dimensions change as the business changes or when you’re exploring novel situations, data mining can be an extremely flexible and powerful complement to OLAP.
Data-mining solutions are perfectly suited for sifting through hundreds of competing and potentially useful dimensions of analysis and associated combinations. All data-mining algorithms have built-in mechanisms that can examine huge numbers of potential patterns in data and reduce the results to a simple summary report. The BI OLAP and data-mining approaches to reporting on data belong together and are synergistic when deployed together. Microsoft recognized this synergy after it released SQL Server 7.0 and began a development program to migrate data-mining capabilities into the SQL Server 2000 release. The most common data-mining techniques are decision trees, neural networks, cluster analysis, and regression. In preparing to release SQL Server 2000 and Commerce Server, Microsoft developed a substantial data-mining infrastructure and core data-mining algorithms to carry out decision-tree and cluster-analysis data-mining tasks. As part of the data-mining infrastructure, Microsoft created the OLE DB for Data Mining specification, an extension of OLE DB for OLAP that defines the data-mining infrastructure and COM interfaces that expose data-mining models and algorithms to data-mining consumers.
The Essay on Data mining techniques
2.1Assuming that data mining techniques are to be used in the following cases, identify whether the task required is supervised or unsupervised learning. a.Supervised-Deciding whether to issue a loan to an applicant based on demographic and financial data (with reference to a database of similar data on prior customers). b.Unsupervised-In an online bookstore, making recommendations to customers ...
OLE DB for Data Mining serves as a standard that external product vendors can use for delivering their data-mining functionality in the Microsoft environment. The Data Mining and Exploration group at Microsoft, which developed the data-mining algorithms in SQL Server 2000, describes the goal of data mining as finding “structure within data.” As defined by the group, structures are revealed through patterns, which are relationships or correlations (co-relations) in data. So the Data Mining and Exploration group has captured the essence of data mining: Correlations produce patterns or associations that show the structure of data. Structure, when placed in a business context, can drive a business modelor even drive the business and improve its effectiveness in the marketplace. The Data Mining and Exploration group model of data mining is to deliver indicators of data structure through extensions of the data query process. Traditionally, you construct a query to retrieve particular information fields from a database and to summarize the fields in a particular fashion. A data-mining query is different from a traditional query in the same way that a data-mining model is different from a traditional database table.
The Essay on Data Mining Process Business Customer
In today s business world, information about the customer is a necessity for a businesses trying to maximize its profits. A new, and important, tool in gaining this knowledge is Data Mining. Data Mining is a set of automated procedures used to find previously unknown patterns and relationships in data. These patterns and relationships, once extracted, can be used to make valid predictions about ...
In a data-mining query, you specify the question that you want to examine (e.g., gross sales or likeliness to respond to a targeted marketing offer), and the data-mining query processor returns to the query station the query results in the form of a structural model that responds to the question. The central object of Microsoft’s data-mining implementation in SQL Server 2000 is the data-mining model. The Data Mining and Exploration group built several query wizards to facilitate the process of creating and interacting with the data-mining model so that end users need no query syntax. The OLE DB for Data Mining specification provides COM interfaces that can be accessed directly from a client application, however, so both end users and third-party applications can access data mining directly through query processing. Although the wizard-driven interface is the primary mechanism for accessing SQL Server 2000’s data-mining query engine, clients and third-party applications can access data-mining models by using an OLE DB command object. After a data-mining model structure is built (either by wizard or directly), it is stored as part of an object hierarchy in the Analysis Services directory.
The patternsor structure within the dataare stored in summary form with dimensions, patterns, and relationships so that the predictive or classification power of the data will persist regardless of what happens to the original row-level data that the model is based on. Fowler, H. W Data mining, London, 2002 Mencken, H.L., Computer guide, London, 2003.
