What is data mining?
Data mining is the practice of searching through big data for useful patterns or trends. It is a knowledge discovery process whereby these huge amounts of data (“big data”) are analyzed from different perspectives and summarized into useful info that will bring about (depending on its purpose) profit or help reduce costs.
Data mining is not new but has been around for a while. In a nutshell, data mining is the use of mathematical algorithms and statistical techniques for predictive analyses.
Let’s go back to the basics for a non-technical explanation of data mining.
To mine minerals means you extract something from the earth. Remember miners searching for mineral resources? They don’t mine for fun. It is a dangerous job. People don’t go into the mining business to play around. They enter because they hope to make a huge profit from the precious metals they will hopefully find.
In the same vein, when marketing or retail companies mine data or engage in data mining, it is because they are hoping to find some useful information that will make them a huge profit. They don’t engage in data mining for fun or just to play around with people’s info. Retailers engage in “know your customer” practices (often called the first law of business).
Through data mining and the use of mathematical algorithms and techniques, consumer trends and buying behavior can be predicted. For example, what an individual will access or most likely search for on a website can be predicted. They want to know what you are buying, what you will most likely buy and when you will buy again. However, are they sure that you will really buy again? There has to be a hook; something that will make you come back.
Why do companies practice data mining?
Many companies engage in data mining today for marketing and other purposes. It is very common in the retail industry. Have you ever received something in the mail like a coupon for baby formula when you didn’t sign up for it or tell the store that you are perhaps, pregnant? In 2012, a Minnesota father found out that his teenage daughter was pregnant before she even told him. How? Something came in the mail from Target- coupons for baby clothes and crib.
At the Financial Times Innovate 2012, Tesco’s CEO Phil Clarke talked about how big data has transformed the company and taken them to another level from shopping by fax to Internet shopping.
Relational databases are mined for specific information.
For example, colleges or institutions of higher education, may search for segmented data or mine data for diversity reasons. Some search for the propensity for conversion and to find out how much effort it will take to turn applicants into students perhaps within specific demographics. In such situations, mining data is deemed a form of strategic resource planning.
Is data mining always successful?
Just like miners may hit a dry hole when mining for minerals, so can you hit a dry hole with data mining and efforts wasted. Not all mining expeditions are successful because of bad data and other factors. However, they learn from that experience and try again.
Mathematical algorithms. Statistical analysis. Predictive patterns. Big data. Snooping. Unauthorized access. Consent. These are keywords currently being analyzed by data privacy experts. Perhaps math can unlock the secret of the universe after all. What do you think?
Image courtesy of sxc