Big Data: Applications run to the forefront of research

In the just-concluded "2012 China Computer Conference", "Big Data" is the highest frequency of the word, almost every speaker, each forum is involved in these words. At the same time, the Chinese Computer Society in the General Assembly period, the formal establishment of a large data expert committee, Chinese Academy of Engineering academician Guojie as director.

What is big data? According to Wikipedia, large data refers to a collection of data that cannot be crawled, managed, or processed by conventional software for a certain amount of time.

In 2011, the publication of Nature magazine said that if large data were to be organized and used more effectively, mankind would be given more opportunities to play a huge role in the advancement of Science and technology to society.

The U.S. government sees big data as "the new oil of the future". In March 2012, the Obama administration unveiled a "Big Data development Program" designed to improve and improve the ability of people to acquire knowledge from massive and complex data, thus speeding up the pace of American inventions in science and engineering and enhancing national security. This is the United States in 1993 announced the "Information superhighway" after another major technology development deployment. 6 Government departments and agencies, such as the National Science Foundation, the Ministry of Defence and the Department of Energy, announced an investment of 200 million U.S. dollars.

In fact, big data has long been around us. We upload or download a lot of video, pictures and text every day through the Internet, all of which produce a lot of data, shopping, social networking sites are the focus of large data. International Data Corporation (IDC) research shows that social networks around the world have 250 million photos uploaded every day.

In 2011, global data totaled 1.8ZB (10 of 21 bytes), of which 75% came from individuals. It has been graphically calculated that the data, if stored with a 9GB DVD, is 260,000 kilometers thick and about two-thirds of the Earth's distance from the moon.

Enterprise applications, however, are the main catalyst for big data. Biology, medicine, astronomy, environment, physics, engineering, economics, Internet and many other fields involving large data processing and application. Guojie the analysis of IBM, Oracle, Microsoft, Google, Amazon, Facebook and other enterprises. He found that since 2005, IBM has invested 16 billion of dollars in 30 major data-related acquisitions, prompting its performance to grow at a steady pace. In 2012, IBM's share price exceeded the 200 dollar mark, doubling 3 times times within 3 years. Wall Street has long been recruiting astronomers and theoretical mathematicians proficient in data analysis to design financial products. IBM is now the world's biggest employer of mathematics, and mathematicians are applying its data analysis to oil exploration, health care and other fields. Ebay uses data mining to accurately calculate the rewards that each keyword in the advertisement brings to the company.

Large data applications are already in the business world. Kingdee International Software group chief Scientist Zhangliangjie introduced, they participate in the construction of the national SME information platform, bringing together 40 million enterprises, through the massive data mining and analysis of these enterprises, can make accurate early warning of economic operation, help the relevant State departments to make decisions.

In addition, Zhangliangjie, for example, in the Microeconomic field, the role of "big data" has become increasingly prominent. "A US company that deals with weather forecast information and data on Amazon's cloud platform can help farmers in agriculture avoid meteorological disasters during certain seasons and protect their earnings well." In enterprise management, large data can bring vigor to enterprise management innovation. Analyze the performance of the previous period to decide the next step of investment, strategic deployment, product development. ”

Domestic Taobao, Tencent QQ, Sina Weibo is not only the production of large data, users, but also researchers. As a result of these applications, Guojie that "large data engineering research has been ahead of science." ”

But when technology cannot solve problems more and more, scientists need to find common problems. However, scientists do not understand the characteristics of each area and problems, the current various fields of "data" in the common is not clear, so Guojie put forward: "I think we engage in science, to put down the body, willing to supporting roles." When conditions are not yet ripe, computer scientists should humbly be "assistants" for a period of time. ”

Hua Yunsheng, a professor at the Chinese University of Hong Kong, said: "The situation is so complicated that, unlike before, a solution can solve many problems." The research of large data should combine the problem of user, solve the problem, data and platform. But it's not clear yet. ”

Guojie that the study of large data has a "subversive" significance for scientists: "For a long time, many fields are using small data to do scientific experiments to find a model and a rule." Now it's getting more complicated, and there's a bunch of data that looks irregular. Scientists need to find new ways of doing this, and this change in patterns and methods needs to be explored. To study the problem of large data, what technical problems found in the application, these targeted problems, not by fantasy, is to be summed up in practice. ”

As for the development of large data, Guojie is optimistic: "The Chinese people, Internet users, the need for large data, strong demand, this is a challenge to the scientific community and opportunities." But Guojie also suggested that much of China's data is now in the hands of different government departments. "Some departments use some data as their own, where they are not used or provided to researchers." Large data needs to be shared, data open and platform utilized, which is the weakness of large data research in China and the big problem to be solved. ”

(Responsible editor: The good of the Legacy)