Explore Big Data theory

The first big data concept was presented by McKinsey, who believes that in today's world, data that has penetrated various sectors and their business functions has become an important reason for production personnel to exploit and use massive amounts of data, and it is clear that new productivity gains and consumer profitability are coming.

The first definition of big data in the industry is IBM, which expands and divides it into four traits: quantity, variety, value, and speed. In-depth analysis, these four levels of big data can be analyzed: first, the amount of data is huge, for the initial measurement of big data units is at least p (equivalent to 1000 t), E (equivalent to 1 million t) or Z (equivalent to 1 billion t), followed by a very rich variety of data types, for example, there are blogs, video , pictures, and location information; Again, the lower the density creates the higher the commercial value; Finally, the processing speed of big data and the traditional DM technology are essentially the difference.

In fact, however, these traits do not really tell you all the features that big data should have, in fact, there are more big data features that we need to discover, such as analytical, social, research, and so on.

As the old saying goes: three points by technology, seven by the data, who gets the data, who the world is. Some may question the validity of this sentence, but in fact no matter who has said it, this sentence is the same truth. Viktor Mayer-schönberger in his book "The Era of Big data" to illustrate the fact that the era of big data has come, so we must use the exploratory thinking of big data analytics to tap into the value of big data itself and the outside world, including its potential value. In his book, he focuses on how Google uses people's history of search to do two of BI data mining to get more value, the most impressive of which is the use of search records to predict influenza infection somewhere. In addition, the author describes how the Amazon website buys the user's historical browsing history data to recommend different kinds of books for specific users, and later results found that doing so could have a dramatic impact on sales revenue. There are also some U.S. ticketing systems that use all the data from the ticket prices in the past 10 years to estimate when it is appropriate to start releasing the tickets, mainly to get a plan for a substantial increase in profitability, and the final result shows that it does have a better effect.

So the question is, how do you decide that a thought is big data? Viktor Mayer-schönberger writes that big data does not exist for sampling, but rather includes samples of all the data, and that it focuses not on accuracy, but on efficiency, and that big data focuses on relevance rather than causal correlation.

Other experts also have some unique ideas about Big data:

The current data is not big, the data becomes really interesting because it is online, and this is the characteristics of the Internet.

The function of a product that is not present in the Internet age must be its value, and the product that exists in the Internet age is the value of this product.

"For more information on business intelligence, business intelligence solutions and business intelligence software downloads, visit Finebi Business Intelligence official website www.finebi.com"

Explore Big Data theory