The big data age requires imagination.

Remember that in the 2005 when the industry preached the WEB2.0 Revolution, there were two types of reactions. One kind can be called the nihility faction, they said that discusses this matter to be illusory, does not have the money immediately the pattern. A class can be called stasis, and they say there is nothing new about discussing it, and that our so-and-so things have long been WEB2.0. The subsequent years of industrial development attest to the ignorance and absurdity of these two claims.

Today, as the industry preaches the big data age, the same two types of responses are still the most common. The nihility Pie said big data Yunshan, see what money way. The stagnation faction says big data has what novel, Lao Tze's data is very big, the excavation technique is very high. I think the industrial development in the next few years will prove the ignorance and absurdity of these two kinds of statements.

Some people who do not want to read something seriously, see the word large data, literally to think of large data, but forget the other basic characteristics of large data, need to be repeatedly reminded.

One of the characteristics of large data is the integrity and comprehensiveness of the data. Many friends in the industry talk about the big data, the habitual site to point out their own stocks, or those can directly from their own services can be obtained. Considering the current development of the Internet at a very early stage, the existing network services are simplified, distorted, one-sided to the real world of enrichment and cutting, the resulting data is messy, fragmented, local, which contains a very limited amount of gold. If you agree that everything in this world can and is being http://www.aliyun.com/zixun/aggregation/14294.html and networked, then the big data of > will necessarily be complete and comprehensive, This includes not only the user behavior data obtained by the network company through its own service, but also the social, economic, political, and natural aspects of the data. These data, of course, are dispersed in the hands of different companies, institutions and government departments, and integration is not easy, but operational difficulties do not negate the integrity and comprehensiveness of the large data itself. Today, when we discuss the arrival of the big data age, because the development of the Internet to the current stage so that the real world of data development to a certain extent, a variety of information terminals to a certain extent, the cost of data access to a certain degree, so that complete and comprehensive data is not only an ideal, is becoming a reality.

The second feature of large data is the openness and commonality of data. It is precisely because the overall large data is difficult to be obtained by a company, agency or government department, so large data must be generated in an open, public network environment. This openness and commonality depends on a number of network open platforms or cloud services, as well as a range of data standards and norms that are legally supported or socially recognized. Any closed or one-way data can not be large data, no matter how large the size of the data.

The third characteristic of large data is the dynamic and timeliness of data. Astrophysics and theoretical physics have long relied on a large amount of data obtained from the universe, and similar disciplines include environmental ecology, medicine, and self-control technology. But that's not the same thing as the big data we're talking about today. Today's big data is based on the Internet's timely dynamic data, not the historical or strict control of the environment generated things.

So the big data we're talking about today is comprehensive, open, dynamic, and timely, and such big data are new challenges that we've never had a chance to take advantage of in the past, and new strategic opportunities that we should strive to harness in the future. If anyone thinks that the data accumulated in the past is large data, or that the past accumulated capacity and experience of data processing can be a natural leader in the big data age, it is not ignorance or arrogance.

There has been a lot of recent media progress in the direction of big data, one of which is a good illustration of the nature of the big data and the prospects for its use. According to the New York Times, August 30, a start-up company called Climate Corporation (Climate Corporation) has more than 10,000 simulations of weather conditions over the next two years in more than 1 million locations in the United States. The company then combined the data on root structure and soil porosity with simulation results to provide crop insurance for thousands of farmers.

The acquisition of soil data through remote sensing is not the same as what we used to acquire user network behavior data through Web services, and the concept of data is greatly expanded. Every day for more than million sites for millions simulation, the volume of data, dynamic, timely. To provide accurate insurance services for each field, it is also necessary to have agricultural futures data matching with land data, climate prediction data, international trade data, international political and military security data, national economy data, industry competition data, etc. The business model, based on such a large and complex data base, is innovative and highly competitive and sustainable and large-scale compared to existing crop insurance practices. Even better, the company, based on large data operations, did not invest heavily in internet facilities, but rented Amazon's public cloud services for one months and tens of thousands of dollars.

If you look carefully, there are a lot of cases like this, although they are relatively simple and rudimentary, but sufficient to illustrate the problem. If we expand our imagination, similar to the above case