Offline Data Mining Strikes Gold
07 June 2010
You'll often hear the term "striking gold" associated with data mining. Just as gold miners received information about a patch of land and went in with their shovels hoping to strike it rich, data mining deals in relatively the same way. The process is becoming popular for businesses of all types, and if done right it can be an extremely low-risk, high-reward process.Basically, data mining is the process of discovering and analyzing data from different perspectives. Once data is compiled and analyzed, it is then summarized into useful information for a business. The result, hopefully, will help to cut overhead costs, increase revenue and be an all-around tool for business improvement. In a sense, you can think of data mining like election polling. With a strong sample group of voters, proper analysis can paint a picture of who's going to win the election. If you'll notice, however, there's a catch in this process. A person (statistic) would have to be present within a field in order to give a result i.e. a voter would need to be polled instead of a random person.
Anything quantifiable is data. You can deal with facts, numbers, text, people, and even statistics on shopping habits. Businesses are pressing the limits of what data is, using operational data like cost, inventory, payroll, accounting and sales; non-operational data like forecast data, macro economic data and industry sales; and even meta-data, which is, essentially, data about the collected data. Any collected information can then be quantified to knowledge, and trends can be discovered and predicted. The goal is to mine the data, analyze it and come up with hard data about consumer buying behaviors, employee behavior, geographical significance, and a number of other usable statistics to help your business grow. Not every business is employing this process on the same scale. While some do collect the data in various forms and use it to their advantage, only the companies serious about data mining actually invest in the processing power and build data warehouses where trends are stored and all data is centralized.
