When "Big data" comes knocking at the door

Drawing: Zhang Fangman

When you have a minute to read this line of text:

Sina has sent 20,000 microblogging, Apple has downloaded 47,000 applications, Taobao has sold 60,000 products, Renren has taken 300,000 visits, Baidu produced 900,000 search queries.

2010 Printed Edition Britannica encyclopedia, 32 volumes, weighing up to 58.5 kg. However, its full content, but also installed a 4G u disk. In view of this, the Encyclopedia Britannica Publishing House announced in March this year, no longer print edition, full digital content.

With the birth of broadband, mobile Internet, IoT, social network and cloud computing, a large data age has inadvertently descended. Not long ago, Feng Xi large data industry park quietly settled in Shaanxi Province Xi-Xian new District, the development of large data industry is "test water."

Global data volumes are growing at a doubling rate every two years

Engaged in advertising culture and creative industry Mr. Ho, clearly remember, from 12 ago, only 20GB hard drive capacity of the home computer, to meet the needs of the continuous expansion of the 80GB, 120GB, 250GB, 500GB host storage space, change quickly. "2TB of hard drives are not available today and are backed up by mobile storage devices. ”

According to the IDC (International Data Company) monitoring statistics, 2011 Global Data volume has reached 1.8ZB (1ZB equals 1 trillion gb,1.8zb is also equivalent to 1.8 billion 1TB of mobile hard disk), and this number is increasing every two years, is expected to 2020 globally will have a total of 35ZB data volume, growth of nearly 20 times times.

Because of the rapid expansion of data scale, the accumulation of data in various industries has become more and more huge, data types are more and more complex, has gone beyond the traditional data management system, processing mode of competence, so "big data" such a meaning to approach the "infinity" concept will emerge.

The first is the concept of a complete collection of data, the National Ministry of Industry and Software services Secretary Albert Chan summed up the four characteristics of large data, "the second is more types, including structured data, semi-structured data, unstructured data and other types, including video data in the current accounting for more than 90% of the total The third requirement is fast, and it needs to be processed in real time with the target of second level. ”

"Finally lies in the value density", Chen Wei thinks, with a lot of useful and potentially useless data coexisting, "Everywhere is gold, and everywhere is sand", so the purpose of large data is to search for valuable data and knowledge from a large collection of data, through analysis and excavation to provide real wisdom for various industries, "can say 21st century is ' Data to drill out of the oil ' era. ”

"Take interactive data for example, at present, some media platforms, such as Sina Weibo, there are more than 25 million micro-blog information published every day, which has a lot of valuable information has not been excavated," China Electronics and Information Industry Development Research Institute deputy Chief engineer June, in such a large number of unstructured data behind, The use of large data technology, from the massive accumulation of interactive data found with a trend, forward-looking information, you can find and produce great social and commercial value.

Big data behind a small apple: the sum of data is worth much more than the value of the data.

"Because of the storage of data, analysis, application and other aspects of business operations have not stereotypes, the development of the industry's potential, innovative space is very large," Shaanxi province, Xi ' an new District Management committee deputy director, Feng XI Xincheng Management Committee director Liu Yubin played a "small apple behind the big data" simple analogy:

Taking the development of Apple industry in Shaanxi Province as an example, the spatial geographic information data needed for the optimum growth of a certain breed of Apple with specific production areas of apple output, sugar content and other data superimposed, and through the Internet and other means to give Apple traceability to the only "identity" in the growing process of real-time monitoring, by each Apple "feedback" The data collected, if sufficient mass, will be integrated into large data.

With this data, you can first generate value through data rental services and potential customers, "this business model embodies the value of the sum of data far greater than the value of the data." ”

Second, if you can use Group analysis, data mining and other scientific methods, supplemented by cloud computing, distributed storage and other means, can carry out in-depth analysis and prediction of data Services, "which Apple is the best quality, better market response, next year, production sales will be what the market for Apple to buy the preferences will be changed", This kind of data deep digging and the consumer behavior forecast analysis behind, once was the statisticians ' privilege, in the future may spend several minutes time can complete.

Data accumulated and compared with other places in the country, it can provide decision support service for the development of Shaanxi Apple industry, and become the decision basis for the Government and industry to guide the production of fruit growers, so as to avoid the unsalable products and the benefit of the farmers.

Finally, with the establishment of the authoritative data and analysis methods, it is possible for the data service providers to build a third party large data analysis platform to provide data collation, filtering, analysis and processing services for more data holders, and even one day develop similar to ebay, Taobao and other E-commerce trading platform, the same Third-party data-sharing trading platform.

Shaanxi to implement large industrial electricity price in large data industry layout, China Unicom project 8,000 cabinet a day to save electricity more than 400,000 yuan

In the West Salty new district planning 25 square kilometers information industrial Park, the first large-scale data processing and service professional park in China--------------------------------------------- Shaanxi Province and strive to 2017, the completion of the West as the core of the national large-scale data processing and service industry cluster, by the end of 2020, the large data industry park will realize the output value of 50 billion yuan, Shaanxi province, Ministry of Industry and Trade department deputy director Cai Suchang said.

Baidu Company for each purchase of a server, the cost of about 30,000 yuan, but the use of maintenance costs more than 30,000 yuan, "in this case we have to continue to ensure that the data center energy-saving consumption." "Baidu Technical Committee chairman Chen Shangyi said.

It is understood that Shaanxi Province has been in the large data industry layout, project aggregation, financial support, infrastructure construction has formulated relevant policies, including the implementation of large industrial enterprises in the industry electricity prices, and exempted from the network fees. "Electricity prices account for about 75% of the cost of data companies, at present, the park has three operators and the national Population Information Processing and backup (xi ' an) center and other projects settled, take China Unicom project 8,000 cabinets as an example, one day can reduce the cost of electricity costs of more than 400,000 yuan, Liu Yubin introduced.

The challenge of big data is not just the "hardware" level. "100 years ago, doctors can understand all the branches of medicine, but today a doctor is facing about 10000 kinds of disease syndrome, 3000 drugs, the vast knowledge of 1100 test methods," said Hequan, deputy director of the National Informatization Advisory Committee, and academician of the Chinese Academy of Engineering, Large data industry needs practitioners to understand both data analysis tools and industry analysis, and such innovative talent is scarce, "at the same time, in the large data mining and development and security and privacy protection, China also lacks the corresponding legal protection, the need for mechanism innovation to promote the realization of data sharing." ”

(Responsible editor: The good of the Legacy)