Wednesday, August 12, 2009

USING CHARTS FOR EFFECTIVE DATA MINING

Data mining is basically a process of identifying underlying patters and relationships among sets of data that are not apparent at first glance. It is a method by which large and unorganized amounts of data are analyzed to find underlying connections which might give the analyzer useful insight into the data being analyzed.


The uses of Data Mining are varied. In marketing it can be used to reach a product to a particular customer. For example, suppose a supermarket while mining through their records notices customers preferring to buy a particular brand of a particular product. The supermarket can then promote that product even further by giving discounts, promotional offers etc. related to that product. A medical researcher analyzing D.N.A strands can and will have to use data mining to find relationships existing among the strands. Apart from bio-informatics, data mining has found applications in several other fields like genetics, pure medicine, engineering, even education.


The Internet is also a domain where data mining is used extensively. The world wide web is a minefield of information. This information needs to be sorted, grouped and analyzed. Data Mining is used extensively here. For example one of the most important aspects of the net is search. Everyday several million people search for information over the world wide web. If each search query is to be stored then extensively large amounts of data will be generated. Data mining can then be used to analyze all of this data and help return better and more direct search results which lead to better usability of the Internet.


Data mining requires advanced techniques to implement. Statistical models, mathematical algorithms or the more modern machine learning methods may be used to sift through tons and tons of data in order to make sense of it all.


Foremost among these is the method of charting. Here data is plotted in the form of charts and graphs. Data visualization, as it is often referred to is a tried and tested technique of data mining. If visually depicted, data easily reveals relationships that would otherwise be hidden. Bar charts, pie charts, line charts, scatter plots, bubble charts etc. provide simple, easy techniques for data mining.


Thus a clear simple truth emerges. In todays world of heavy load data, mining it is necessary. And charts and graphs are one of the surest methods of data mining. And if current trends are anything to go by the importance of data mining cannot be undermined in any way in the near future.