Big future of Big data

"We think we need information, but actually what we really need is knowledge." ”

--"signal and noise"

| Wen Ben journalist Cai

As early as 2008, the famous Science journal Nature introduced a cover article called "Big Data" detailing the increasingly important role of "data" in mathematics, physics, biology, engineering, and social economics. Now this obscure phrase has quietly become the new darling of business and finance. Conferences and forums on big data have sprung up, but what is big data is still a lot of controversy. "We believe that large data should have large, the four characteristics of high value, cross multiplexing, and holographic visibility, especially the last two concepts, reflect the progress of large data not only on the amount of ' larger data ', but also on the qualitative changes that are different from previous data organization and application forms. "Talking about the prospect of large data, Beijing Yong Hong Shang Zhi Technology Co., Ltd. General manager Arlene appears unusually excited.

It is understood that the general view that large data is the forefront of data analysis technology, is a new generation of information technology and industrial development of the focus area, with the development of modern market economy, the large data will become the global overall development of the engine trend is increasingly obvious. It is foreseen that the "big Data Age" will come sooner or later, the European and American developed countries to the data assets protection and use of the rise for the national strategy. On the other hand, according to IDC, China is poised to open the world's largest data market soon, with a market size of more than 2 trillion yuan in 2020.

Ferocious large data

"If every single data is isolated, only in the areas directly related to his own value, it is not a data revolution era, we have to find the implementation of data between one and a far greater than the value of two, during which the most critical problem is to play the external data. "For new forms of organization and application between several data, Arlene also has his own ideas," he told the new era of China reporter, with the popularity of mobile internet and statistical methods are improving, large data is rapidly expanding its application. "For example, the state grid through the smart meter system can estimate housing vacancy rate, Alibaba through Taobao sales data can judge the trend of economic operation, Micro-Bo ax concern and content information can be used in advertising push and so on." ”

"Specifically to Yong Hong Technology, because of the size of the reasons, we can not be compared with those technology giants, but this does not affect the rapid development of the company." In the past 2013, we have ushered in a large amount of venture capital, although there is no small gap with the Giants, but in our field has been good. Arlene added.

But other companies are not so fortunate, according to Arlene, even if the big data in China to develop quickly, but every year there are still many companies fail, the biggest reason is the direction of the lost. "Take the user as the center, unifies the user in the different domain to leave the track, realizes the data intersection, this is the big data produces the value the most feasible way." "Arlene said.

At the beginning of 2012, Beijing Yong Hong Shang Zhi Technology Co., Ltd. was founded at Beijing Institute of Technology. As a High-tech enterprise in Beijing, Yong Hong Technology is mainly committed to radio and television, telecommunications, mobile, security, Internet, government and other large-scale industry professional advisory services and system construction, and accumulated a wealth of data resources. Although the results are good, but Arlene for the company's main business is not at ease, with his words that is, "some people happy people worry, ferocious big data like unstoppable wave, swept all of us, the remainder of the king, to live to be bigger." ”

The beginning of all this, in the final analysis depends on the development of the Internet industry. It is the development of the information generation, organization and circulation of the way has revolutionized the change, and the most important change is the first single individual to become the main body of information generation and circulation. The average person uploads a photo of the network, a video, emails sent, chat software chats, shopping on e-commerce sites, paying with credit cards ... all internet-related information flows are converted into data and stored in servers. In just more than 10 years, many enterprises climbed the MB era, through the GB era, is now being driven to run the TB era, to meet the PB era. In the present China, many big companies only daily update data volume is already near or reached the PB level.

But there are also different views that, while the size of the data is exploding, the added value of the data does not seem to be synchronized with it, and even some scholars believe that the density of the data will decrease as the amount of data increases. Arlene also told the Chinese New Age reporter his view, "I think there is no absolute causal link, the crux of this lag is the lack of value from the vast number of data mining efficient methods and technical personnel, which is the direction of our efforts." ”

He told reporters that if the value of a group of data is the natural logarithm of its size, when you earn 9 dollars from 1GB of data, give you 1PB of data, you can only earn 15 yuan. If the value of the data can be fully mined to make the value of the data proportional to its size, then 1PB of data can bring the benefit is 9 million yuan. "This is a huge difference, so for real big data, the growth in value should be proportional to the growth of the scale, even faster than it is." "said Arlene.

Recall a few years ago the hardships of entrepreneurship, Arlene Frankly, positioning precision is the key to the successful development of Yong-hong Science and technology. It is understood that Yong Hong Technology management is a young team. As the national key support of international students entrepreneurial enterprises, the company's management members have overseas study and experience. "Our first big data customers came from the telecoms industry, and they found us, when they didn't have very good customers and projects, and then they went to other international and domestic vendors, and we thought we might not be expecting much." But after one months they came back, decided to cooperate with us, and later he and we said: inspected a lap, can not find the Bilonghon technology more suitable manufacturers. Slowly, the company's customers more up, the next step we intend to visit Guizhou, where to look for new opportunities. "said a co-founder of Yonghong technology.

Set sail in Guizhou

Yonghong Technology expects Guizhou as its starting point for two times, not without reason. Although Beijing and other big cities of talent, market advantages are very obvious, but such as policy, geographical agglomeration, human costs, such as is unable to make up the short board, and this is the advantages of Guizhou.

At the end of 2013, the spring Breeze, which was approved by the State Council in Guiyang Hi-tech Development zone, became the center of China's large data industry. Almost at the same time, a large number of data giants such as Chinese Telecom, China Mobile and Unicom (600050, shares) have broken ground in the new area. It is understood that China Telecom Cloud Computing Center covers an area of 500 acres, the total investment of 7 billion yuan, the first phase of the completion of the server capacity of 1 million units, from the end of 2014 can be put into commercial. China Mobile (Guizhou) Data Center project total investment of 2 billion yuan, the land of about 275 acres, the total project size of about 210,000 square meters, including 160,000 square meters of new green data center room, 10,000 square meters storage room, 40,000 square meters production support room. China Unicom (Your) cloud computing base covers an area of about 500 acres, plans to invest about 5 billion yuan, the main construction infrastructure, data center resources, disaster preparedness system, computer room construction and other facilities.

Deming, director of Guizhou Communications Administration, has previously said that with China's "Four synchronization", "Two fusion", "Industrial Upgrading" and a series of industrial information support policies, information has become an integral part of the development of Guizhou. As the physical carrier of the industry application informatization, Guizhou Data Center is bound to have more and more enterprises, especially the communication operators to start new or upgrade their own data centers to improve their informatization level and enterprise core competitiveness. There are divergent views on the big data in Guizhou, but nothing more than one, we mainly do is the Internet server hosting and management, these are the basic network of Internet infrastructure resources, like backbone network, access network, our goal is to provide Internet users with high-end data storage, processing and transport services. "Deming said.

Arlene is also very optimistic about the future development of Guizhou, "after the completion of the large data base will achieve High-tech, low-cost, green energy-saving." I believe that the new zone can take the development of China Telecom, China Mobile, Chinese Unicom data center and other major projects, develop in data collection, analysis and mining, application services, and promote the e-government, E-commerce, smart tourism, intelligent transportation, intelligence city and other industries Take-off, and finally reached the leading domestic level. In view of the geopolitical conditions, Guizhou, which is about to enter the "high-speed railway era", is an important land transportation hub in southwest China because it is located in the China-ASEAN Free Trade Area and the economic corridor between Burma, India and Bangladesh. In the future, with the policy support of the central and various ministries and the new district of the new district, there are many policies and reforms to be tried first.

Other industry analysts also said to reporters that Guizhou's development of large data industry advantages mainly lies in: Guizhou away from the seismic zone, with a stable geological structure, low risk of disasters, large data enterprises choose to settle in Guizhou can minimize the operational risk of enterprises; At the same time, Guizhou's energy advantages can provide large data enterprises with cheap, Stable power resources, information flow and energy flow can also be better combined; Finally, Guizhou is the base of three military enterprises, brought together a group of aerospace Electronics (600879, shares bar) equipment enterprises and Everest technology and other emerging electronic information enterprises, with a certain supporting capacity of industry supporting.

Re-discovery of large data

"Since the printing press, our world has gone through a lot. Information is no longer so rare, we have too much information, even to the impossible, but only a few. Our instinct is to look at information subjectively and selectively, but it is not enough to distort information. We think we need information, but actually what we really need is knowledge. "The famous American information expert, Natte Silver, wrote in his monograph," Signal and noise. "

In this book, he presents a famous question about whether data is a useful signal for our future or meaningless noises. Silver was the winner of Barack Obama's campaign when the U.S. election was a stalemate, but silver used baseball's statistical methods to draw the conclusion that Obama would win. The prediction gave him a big name. But he soon began to rethink that the case for successful predictions using big data was far from being a failure. For the large data concept now commonly mentioned, the systematic judgment relies on the historical data, but the person relies on the field judgment in addition to the history, sometimes the human potential cannot use the data performance. When silver was using large data to predict the results at the baseball stadium, he found that the players ' faith and focus were the key to making him play well on the pitch. These two points, are not aware of the data system, nor is it in line with the famous "fast, inclusive high" large data definition.

Arlene also believes that many of the so-called Big data definitions are not accurate, or even vulgar the meaning of large data. "From my experience, processing speed is definitely not a feature of large data, but only the demand for Internet Information services." Ten years ago, no big data, but Internet users will not be bitter for one hours, although the amount of information at that time is small, but real-time computing is not as difficult as the current small. "According to Arlene, now a lot of typical large data operations, short to a few hours, long can reach a few months or even years, may have the same great value." "Obviously, 1 seconds to figure out that the result is not a big data feature, the faster the better, from the human invention calculation has begun to explore, as the current data age label, obviously not appropriate." "Arlene said.

In fact, in Arlene's view, the ability to analyze data is the most prominent problem of large data. "To give an extreme example, if Google were to open up to a research or entrepreneurial team free of charge for more than 1PB of data per day, it would be very likely that the data would not help the team's development because they did not have the ability to retrieve, crawl, compute, and analyze the data for that magnitude. Perhaps they are only interested in a particular logical fragment within the data, but they have no way of knowing where the logical fragment is located in the data and how to get it. ”

One side is a rapidly growing amount of data, one is a market that needs data, who can connect both, who is likely to lead in the competition, and this is the Yonghong technology has been operating the "killer." According to Arlene, from the beginning of the establishment, Yong Hong Technology cooperation partners in the electronic data cube, radio and television viewing behavior analysis, large power data analysis, traffic data analysis and other data-link areas of in-depth cooperation. "Large data scale can be very large, but it should be as simple as operating a ' small data ', which requires data to be summed up very well, should be done within a variety of content and relevance is clearly visible and easy to call access." "said Arlene.