How to solve the hot demand of "big data"

The large amount of complex data provides the information foundation for us to understand and grasp the inner law of the development of things more deeply and accurately, which contains great potential value. This is not only the consensus of many industry research institutions, but also has been reflected in commercial applications.


at present, large data applications have a number of typical cases, such as IBM Japan's use of Internet search data set up by the economic indicators forecasting system, and the American University of Indiana using Google's mood analysis tools to predict the changes in the Dow Jones Industrial Index, has reached a relatively high accuracy. In addition to economic analysis, in agriculture, medicine and health, manufacturing and other fields, there are some successful application of large data to predict the case. According to McKinsey, the big data will bring a potential increase of $300 billion trillion a year for the U.S. medical services industry, with a potential value of € 250 billion per year for public administration in Europe, and a potential annual income of $600 billion for the location-based services industry. The retailer can make full use of large data to realize the increase of operating profit by 60%, and the manufacturing industry can reduce the cost of equipment assembly by 50%. A new OECD study also estimates the market value of Internet data, supporting the huge potential value of large data.


relatively, at present, China's large data industry is still at the initial stage of development, the market size is still relatively small, 2012 only 450 million yuan, and leading manufacturers are still in the majority of foreign companies. It is predicted that 2016 China's large data application of the overall market size will break through tens of billions of dollars, the future will form the world's largest large data industry belt. However, compared with the optimistic forecast of the development prospect, the realistic challenge of developing the big data industry in our country deserves to be analyzed and treated seriously.


one is how to make the data dispersed in different departments and subjects be used reasonably and effectively. You need to be clear about which data can be used and which data cannot be commercially used arbitrarily. The second is how to build an IT infrastructure that supports large data. This involves the transformation and utilization of traditional data centers, the construction of new cloud computing storage and processing systems, and how to quickly build a high-speed, accessible Internet access. Third, how to master large data mining technology and training large data professionals. Large data analysis requires the support of relevant technologies and talents. Some IT companies in developed countries have mastered some key technologies of data analysis and prediction, and the technical capability of Chinese enterprises is still relatively backward. McKinsey forecasts that the United States needs large data analysis by 2018, with a talent gap of more than 100,000, more than 400,000. China has just introduced the concept of large data, the problem of talent shortage is more prominent. Four is how to avoid the future of large data industry redundant construction and even overcapacity problem. How to avoid the recurrence of these problems caused by improper government intervention is a challenge to the industrial development system under the current regional competition in China for the orthometric of heat.


in short, the face of the community's "Big data" hot, should be rational analysis, calm observation, a solid do several aspects of the basic work.


first, there is no need to rush to introduce strategic planning and set up industrial special funds. Domestic it enterprises and local governments have been aware of the development prospects of large data industry, the development of large data applications have a greater enthusiasm. Some cities have launched large data development strategies, and are planning to form at least 50 billion yuan by 2017. In such cases, incentives, such as planning and special funding, can distort normal market behaviour and even generate bubbles.


second, reasonable transformation, construction and layout of IT infrastructure. For the existing traditional data center and a large number of old server resources, can be established through the establishment of virtual data center or the near-merge, such as the transformation and utilization, explore how to use virtualization technology and cloud computing platform management software to improve efficiency. The new large-scale cloud computing data center should be integrated, reasonable layout, coordination between different provinces and cities to strengthen complementary cooperation, energy and climate factors as the important condition of project construction, to ensure economic rationality. According to the principle of the network construction moderately ahead of the industrial development, speed up the speed of "broadband popularization speed-increasing project", solve the network bandwidth bottleneck of large data application development as soon as possible.


third, funding large data competition before the technology research, training large data analysis talent. In the existing government science and technology projects, the appropriate arrangement of project funds to support the development of key data technology, focus on the pre-competition technology, encourage enterprises to lead or participate in the commitment. The training of talents should start from two aspects of higher education and enterprise technicians, allow universities to set up large data-related majors and recruit students, and encourage local governments to introduce relevant policies on training of large data technology personnel.