Data Mining: From Moore’s Law to One Sale a Day

Today the internet is more customized than it ever has been before. This is largely because of data mining, which involves using patterns and records of how you use the internet, to anticipate how you will continue to use the internet. This is an application of data mining, however; more broadly, the term refers to how to analyze data to cut costs or increase revenue.

While the term data mining is new, the practice is not. Due to Moore’s Law, which states that processing power and data storage double every 18 months, over the past five years, it has become significantly easier to access vast stores of data. People are also continuing to use the internet and explore the web at an exponential rate so that the effect of data mining by 2020 will mean that roughly five billion of the world’s seven and a half billion people will be affected. After about 2020, integrate circuits will be so advanced and tiny, that many predict Moore’s law will be inapplicable to circuitry, but will continue to dictate the conventions of nanotechnology and biochips.

Data mining has more practical examples, too. The products you’ve bought off Amazon, for example, are analyzed by data miners at that company, to show you similar products that you may be interested in. Applied more widely, a restaurant chain could determine what customers buy and when they visit in order to tailor their menu to fit the tastes of the public at large, as well as to invent and supply new dishes and offer specials. This is called class data mining. A deal of the day site could target its giveaway of the day to a certain segment of the population that visits its site. If it knows that most people visit its site searching for technology-related items, chances are it will offer more of those items instead of a clothing or travel deal of the day. This is called cluster data mining. Association mining is a logical rule followed by supermarkets such that if a customer buys bread and butter, he will is likely to also buy milk.

Data mining involves statistics which determine what customers will buy over the course of thousands and millions of interactions. In effect, this is what makes technology seem smarter. The logical and statistical formulae humans implement make these rules widely applicable and largely sensible. The applications of data mining are various and exciting. In the future, the internet will be that much closer to reading your mind.

Post Comment