Data Mining Technologies - Essay Example

It is not possible to give specific advice, but there are four general principles that generally apply: every application must deliver benefits to the users; users must feel comfortable with the way in which the system manipulates and presents information. Data mining is based on statistical analysis and modeling techniques.

Data mining becomes a strategic weapon of organizations because it increases significantly the volume of information that can be stored and the ease with which it can be updated. This enables new services to be provided in a way that was not economically feasible before. The three main steps in data mining are discovering, predicting modeling, and forensic analysis. “Discovery is the process of looking in a database to find hidden patterns without a predetermined idea or hypothesis about what the patterns may be” (Pushpa 2007, p.1). Discovery consists of three subcategories: conditional logic, affinities, associations, trends, and associations. “In predictive modeling patterns discovered from the database are used to predict the future” (Pushpa, 2007, p. 2). Predicting modeling involves outcomes prediction and forecasting. Most of the data mining applications used by an organization consist of simply taking information in one form, performing some very simple transformations to that information (rarely more than aggregation and the usual arithmetical operations of addition, subtraction, multiplication, and division), and then presenting the processed information in another form. Thus, forensic analysis allows both deviation detection and link analysis. It is defined as: ”the process of applying the extracted patterns to find anomalous or unusual data elements” (Pushpa, 2007, p.2). The underlying architecture of data warehousing that an organization needs to best perform these tasks is obvious. There must be some form of flexible, central 'data store' containing all the information and data that are going to be manipulated, and fed by a variety of information-gathering processes. The main data mining techniques are classification, association, and sequence analysis.

Three potential pitfalls in data mining are the selection of a wrong technique and strategy, unclear objectives and aims of data mining, “insufficient number of attributes captured about the customer” or incomplete data (8). The selection of data mining techniques is the core of the process and the main factor of success. Inadequate or inappropriate data mining techniques will lead to wrong results and mistakes. Unclear connectives prevent the company from effective information gathering and will result in financial loss. An insufficient number of attributes and incomplete data will make it impossible to the company to analyze the data and forecast.

The airline industry uses data mining for several purposes: it tries to identify hidden trends in data which would help to improve its services and customer communication. Data mining allows airline companies to gather information about customers of frequent flier program and their travel, and then identify the main trends in customer behavior. This approach can also be used by virtual retail organizations that want to sell as well as buy. This approach will help retail companies to analyze changing behavior patterns and identify new preferences and needs of potential consumers. This method of selling is now being investigated by other industries as well. Data mining proposes great opportunities for a virtual retail businesses to faster their processes and deliver high-quality services to customers. As the price/performance ratio improves, it is not only leading-edge technology that becomes cheaper — so do technologies that were expensive some years ago, and this means that these can now also be exploited in new ways.