Big data: Seize the opportunity to preserve value

Today, data is more rooted in every corner of our lives than ever before. We try to use data to solve problems, improve welfare, and promote new economic prosperity. With the improvement of data processing capability, the blowout of operation and storage cost and more and more devices embedding various sensing technologies, data collection, storage and analysis are in an almost infinite upward trend. In the 2011, it was estimated that people around the world created and replicated more than 1.8 kilobytes of information a year, and by 2013 the figure had risen to 4 kilobytes.

Each day, the world will upload more than 500 million images, and 20 hours of video per minute will be shared. However, even all the information that people create every day--including voice calls, emails and information--and all the pictures, videos and music that are uploaded will not be able to count the amount of digital data that is created every day on people's own.

This trend will continue. We are now at the very beginning of the so-called "IoT of things," and as technology matures, our equipment, transportation and rapidly evolving "wearable" technologies will be able to connect and communicate with each other. Advances in technology have reduced the cost of creating, capturing and managing information to one-sixth per cent in 2005 years, and since 2005 business investment in hardware, software, talent and services has grown by a full 50% to $400 billion trillion.

There are many definitions of big data, and this is a different definition for computer scientists, financial analysts, or entrepreneurs trying to win venture capital. Most definitions reflect a technical ability to capture, sum, and analyze larger, faster, and more diverse data. In other words, "people can now get the data faster, the data covers a broader area, and the magnitude of the data is not what it used to be." "More accurately, a large dataset is a decentralized dataset that is" produced by devices, sensors, network transactions, e-mail, video, click Streams, and all possible digital resources today and in the future, larger, more diverse, more complex and more in-depth. ”

No matter how we define large data as a technological phenomenon, the limitless potential of large data analysis challenges our legal, ethical and social norms to test our ability to protect privacy and other values in the world of large data. The unprecedented computational power brings us incredible discoveries, innovations, and advances in the quality of our lives. However, most ordinary consumers find it difficult to see this power, so it also creates the power between the person who grasps the data and the person who knowingly provides the data. So what's more important than what "big data is" is what "big data can do" and the relationship between individuals and data collectors.

So what are the highlights of big data? The following will define what is truly innovative and truly different from big data. To study large data, we need to pay attention to those number of large, wide variety, transmission speed so that the traditional data capture analysis method is not competent data, such characteristics are agreed to be called "Three V" (Scale volume, species varity, speed velocity). The ever-decreasing cost of collecting, storing and processing data, as well as the emergence of more and more new data sources such as sensors, cameras, and positioning devices, make us now in a world where data collection is ubiquitous. Data collection and processing volume is unprecedented. Networked devices, wearable technology and advanced sensors monitor everything from signs of life to energy use to slow runners, such data explosions have higher computational power requirements and will drive the development of more complex data management technologies.

Data is more than just more, and its origins are broader and more diverse in form. Some data are "digital native", which means that it is a digital information created by a computer or data processing system, such as email, web browsing, and GPS location data. Other data are "virtual native", meaning that it comes from the real world, but is artificially converted into digital form, such as audio and video from mobile phones, cameras and video recorders, or physical activity data such as heart rate and perspiration, which can be monitored by wearable devices. As data fusion capabilities increase, more and more disparate data are brought together, and large data will give people extraordinary insights.

In addition, the speed of data collection and analysis is gradually becoming real time, which means that data analysis can have an immediate impact on people's environment, or even the choice of people. A typical example of high speed data is to record the user's click Stream data, the GPS real-time positioning data on the mobile device, and the user sharing on various social media. Both customers and companies have greater real-time analysis and real-time benefit requirements. Of course, a mobile phone map application if not real-time, accurate positioning, obviously lost its meaning; the real-time operation of vehicle system is also the key to ensure our driving safety.

If applied properly, large data analysis can improve economic efficiency, improve consumption habits, improve government services, and even save people's lives. Like what:

The development of large data and "Internet of things" has the potential of integrating industrial economy with information economy. Today, we can equip sensors on aircraft engines and freight cars, monitor various data, and issue automatic alarms when repairs are needed. Not only can reduce maintenance costs, but also enhance security.

The Medicare and Medicaid centers have started using predictive analysis software to mark potential health insurance scams, the fraud prevention system, which helps to identify high-risk populations in real time, has been successful in preventing fraudulent withdrawals involving 115 million of dollars, and in the first year of applying this technology, $3 trillion is earned for every 1 dollars invested.

Three or one large data studies have compiled millions of data samples from the neonatal intensive care unit to determine which newborns are more likely to receive fatal infections. By analyzing the data, the system can identify early signs of many infections, such as elevated temperatures, heart rate acceleration, and so on, without the attention of the doctor on duty. In traditional practice, some early signs are difficult for seasoned physicians to notice.

Large data technologies also have limitless prospects in many other areas, such as managing power grids on demand, improving energy efficiency, increasing agricultural production in China, preventing the spread of communicable diseases, and so on.

New opportunities also pose new challenges. Large data technology has gained value from huge data sets, giving the researchers unimaginable insights. Of course, the technical power of large data is so broad and far-reaching that we also need to pay attention to the challenges it poses to social ethics. Melvin Kranzberg, historian Kranzborg, wrote in his first law of technology: "Science and technology are not good or evil, but they are not neutral." "We should always keep in mind that technology can be both public and human." The key is to find the best balance between opportunities and challenges.

(Responsible editor: Lu Guang)