by Max Gallo

Created on: July 01, 2007

Data mining, as the term implies, requires first of all data - usually lots of it - and then an action "the mining" of the data. In its simplest form it is simply that; sifting through large volumes of data to find meaningful information - the meaningful information can be quite different for each "miner" as each miner will set different grouping, clustering and/or selection criteria to make sense of the data.

Where does this data "live". Well, data is collected by hugely different means, organizations and systems every day; credit card purchases, what is purchased, when, how often, where, etc. Electronic tags on cars, where it travels, when, how often etc. You name it and you will probably find it has data collected about it.

This is not to imply that every bit of data that is collected is then mined, but potentially it could be. That is why there are laws governing the sharing of sensitive data among organizations. But, potentially data mining could be done on every single bit of data about an individual that is electronically stored somewhere.

In electronic data mining it is then not surprising that there are things such as Data Warehouses and software that allows the manipulation and mining of the data in these warehouses by authorized - we hope - entities.

Data mining allows for fast and accurate views of data that reveal patterns which would otherwise go unnoticed because of the volume of the data impossible to humanly process and/or for the diversity of the connections which would in small quantities appear to be unrelated. Data mining is especially useful to governments, tax agencies, policing, and other organizations with "high" access authority. It is also vastly used by multi-nationals, shopping chains, manufactures and almost any company that has data - that means any company today; who hasn't got data.

The Internet, more specifically Internet search engines are a sort of data mining. They collect - automatically - according to some algorithm information about all the visible/accessible pages on the Internet, then sort and group them in order to allow for meaningful searches.

So, if the data is accessible/visible to the miner the miner may get information from that data then use that information in an "intelligent way" and perhaps create knowledge. The negative cycle could end up with using information for inappropriate purposes - a form of knowledge too I guess.

Learn more about this author, Max Gallo.
Click here to send this author comments or questions.