Abstract
The process of taking business decisions is taking a more scientific approach with the increasing use of information obtained from hidden patterns in data. Data mining entails the process of extracting useful patterns from large datasets to guide decision taking.
The steps involved in the decision process include data selection, data cleaning, data mining and evaluation. Owing to the size of the dataset involved in the data mining process, algorithms that identify patterns have to be optimized.
A number of tools to make the process of decision support with data mining easier have been developed as the trend increases. One such application of data mining is the identification of consumer behavior through patterns obtained from consumer purchases information.
Business decisions are now being made with quite a lot of intelligence and precision hinged on finding useful information that would support decisions based on available data. Data mining is thus the activity which entails finding trends or patterns that are inherent in large datasets so as to guide future decisions (Ramakrishnan and Gehrke, 1997, p 707). This application of knowledge in decision making is rapidly growing and so a number of tools that support decision are now available to make the task of knowledge discovery from data easier.
Although the process of querying a database can also be thought of as looking for trends, data mining uses higher level querying based on multi dimensional data model and the data set involved is quite large. Specialized algorithms are designed purposely to optimize the process of pattern searching through large datasets.
The knowledge discovery process in the mining of data from a large dataset can be broadly broken down into four steps which are stated as follows:
Data selection: this process involves the identification of the target dataset and the relevant attributes from the raw data to be used for the knowledge discovery.
Data cleaning: because datasets often contain a lot of noise (unwanted information), the data cleaning process removes noise, transform some fields and generate new ones, denormalize relations etc.
Data mining: the process where the actual patterns are extracted
Evaluation: this is the last step in the knowledge discovery process where the patterns identified are presented in a form that can be easily understood by the user.
Using data mining, information on the behavioural tendencies of consumers can be obtained from the data containing the transaction information of consumers that have been collected over a long period of time. This information will be very useful in making decisions on product stock acquisition and arrangement in stores.
WORKS CITED
1. Ramakrishnan, R. and Gehrke, J. (1997). Database Management Systems (2nd ed.). McGraw-Hill.
