Large data age leads to more rational and reliable decisions

Victor Maire-Schoenberg, a professor of internet research at the University of Oxford, and author of the Great Data Age: a more rational and reliable decision in the big data age

What is the magic that makes the concept of "big data" a universal concern for all countries around the world? What exactly is "big data"? How far can it change our lives? As we seek answers to these important questions, Victor Maire-Schoenberg, professor of Internet research at the University of Oxford's Cyber Institute, is now in our sights, and hopefully our interview with him will help readers find answers to these questions.

In recent times, the "Big Data" boom has swept the world, as Forbes magazine has said, and it is almost impossible to see or hear the word "big data" when browsing news sites or attending industry conferences. Last year, 6 federal agencies in the United States announced that they would launch a "Big Data development Program" to invest more than 200 million dollars to improve their ability to acquire knowledge from massive and complex data. At the same time, China's Ministry of Science and Technology issued the "' Twelve-Five ' national Science and Technology Planning information Technology Field 2013," A guide to the choice of projects, also put large data research in the first place. At the current national "two sessions", the NPC representatives proposed to the development of "big data" to the national strategy.

What is the magic that makes the concept of "big data" a universal concern for all countries around the world? What exactly is "big data"? How far can it change our lives? is the focus on "big data" too high? As we seek answers to these important questions, Victor Maire-Schoenberg (Viktor Mayer-schonberger), a professor at the Internet Institute of Oxford University, is appearing in our vision to discuss "big data", if he is not the most suitable candidate, Or at least one of the right candidates.

For more than more than 20 years, Victor has devoted himself to the research of network economy, information and innovation, information supervision, network regulation and strategic management. Victor has conducted a systematic and in-depth study of the concept of "big data", and in 2010 he published 14 pages of large data articles, together with Data editor Kenneth Couqueil in the British Journal of Economics. It is not too much to call him one of the first data scientists to have an insight into the trends of the big data age.

The economist said that in the big data field, Victor is one of the most respected omni-directional spokesmen, the U.S. science magazine said, to launch a discussion on this issue, there is no better than his initiator.

In addition to theoretical research, Victor is very close to the actual world, as early as college, he has set up two data security and anti-virus software companies, and in his book "The Big Data Age," the most cutting-edge, the most new large data application cases, has benefited from his years of keeping pace with business and business applications. Among his consulting clients are large data pioneers such as Microsoft, Hewlett-Packard, IBM, Amazon, Facebook, Twitter, and Visa.

At present, Victor is also the EU's Internet official policy behind the important makers and participants, especially, he also served in Singapore Ministry of Commerce, Brunei Defense Department, Kuwait Ministry of Commerce and other departments, particularly familiar with the Asian Information industry development and strategic layout.

Hopefully our interview with Victor via email will help readers find answers to these questions.

Lose the microscopic level of precision in order to gain insights at the macro level

Wen Wei Po: Today, "Big data" has become a hot word in the world, when did you start to pay attention to it?

Meyer-Schoenberg: Over the years, I have been working on the important role of data in the development of the information economy, and I have published a series of related studies with Kenneth Couqueil (Kenneth Cukier, my co-author). About three years ago, at a meeting of my own organization, we both realized that the existence of "big data" was not just a hype or a grand manifesto, but that it would actually change our work, our lives and society as a whole, so we decided to write a book on the subject.

Wen Wei Po: So in your opinion, what is the big data age? What is the difference between it and the traditional data age? We know that companies like Wal-Mart have used big data in business practices long ago.

Meyer-Schoenberg: In fact, over the past few centuries, the data have played a role in the decision-making process of scientists, and in the past few decades this has extended to the decision-making process of some companies. But before the big data age, data was scarce, and we had very little data. Therefore, our decisions, the systems we build, are based on such a lack of data. Today, everything is very different, it is embodied in three different aspects, we call it "more", "more chaotic" and "relevance".

Wen Wei Po: These three features are also highlighted in your book "The Big Data Age," and they can even subvert the whole way of thinking of our past. Can you describe exactly what this process is?

Meyer-Schoenberg: OK. What I mean by "more" is that around any particular problem that we want to investigate, or questions that require our answers, we can all get more data than ever before. In the big data age, we can use a lot of data to get very detailed insights, which the traditional method can not do.

It can be said that the difference between the big data age and the traditional data age is like an old digital photograph with a resolution of 2 million pixels, which is raised to 24 million pixels. The latter is a very very large file and it can provide more details. It allows us to zoom in and see the small to granular detail, while images with lower resolution are very blurry in these details.

Genetic information is a good example. A new company called 23andMe in the United States provides personal DNA test analysis to detect signs of disease. It costs only two hundred or three hundred dollars and alerts clients to personal quirks that develop into serious illnesses. But the company does not sequence the whole genome of each client, but rather the sites of known features (the DNA fragments that may have been found to be problematic because of a disease). This means that when a new feature is discovered, 23andMe will have to again sequencing the client's DNA and creating a more complete file.

Apple's Steve Jobs tried a very different approach. After he had cancer, he had his entire genetic code, billions of of base pairs sequenced. That cost him more than 100,000 dollars, but it would give doctors a complete insight into his genetic code. Whenever drugs lose their effectiveness due to jobs ' cancer lesions, they can find effective alternatives based on jobs ' specific genetic information. Unfortunately, this did not save jobs, but the data obtained in the process have extended his life.

As a result of technological innovation, the cost of collecting large amounts of information is now getting lower. A few years ago, Steve Jobs spent six-digit sums to do things, and today, less than 1000 dollars will be able to get the same service.

and "more messy" means that in the small data age, because the data is so scarce, we can ensure that every data point we collect is very accurate. In contrast, large data is often messy and uneven in quality. However, compared to the high cost of ensuring the accuracy of measuring and collecting small amounts of data, we will accept this clutter in the big data age, because what we usually need is a general direction rather than an effort to understand the minutiae of a phenomenon. We are not going to give up the accuracy altogether, we just give up our enthusiasm for accuracy. We lose the microscopic level of precision in order to gain insights at the macro level.

Computer translation is one example. At the age of 1990, IBM researchers used a very precise set of documents (Canadian Parliament records in French and English) to train computers. The quality of translation based on this is very low, although the computer is entirely in accordance with the rules. Then, Google began to intervene in the field in 2006, and instead of using the millions of standard translations from the Canadian government, it used any language that was readily available. They use billions of of pages of uneven translations across the Internet, and these translations are not standard-but this is a small trade-off-the data they can use is greatly increased and the quality of translation is improved. More clutter is being won than less and more standard data.

The combination of "more" and "more chaos" creates a third feature, "Relevance", which is also the most fundamental shift in the big data. Our thinking shifts from causality to correlation. Throughout human history so far, people all over the world are looking for the cause of the event and exploring the "why". But our persistent exploration of reason often leads us in the wrong direction. So, we suggest that in the big data age, in many cases, we can just look for "what is" without having to fully understand why. For example, in the analysis of large data, we can find some very small changes in machine vibrations that indicate that the machine will soon be damaged. This allows us to replace parts of the machine parts before they are scrapped, which is called predictive maintenance, which can save a lot of money. But in addition to increasing consumption efficiency, "relevance" can do more.

For premature babies, for example, even if they grow up, these babies are still very vulnerable, even if they are caught in a small infection. Doctor Caroline McGregor studies how to give these babies the best chance of survival. Using large data analysis, which can collect more than 1000 data points per minute, McGregor finds a shocking fact: whenever these premature babies appear to have a very stable sign, their bodies are not stable and are preparing for disease. With this knowledge, she will be able to determine in a very early stage whether a baby needs medication to save more children's lives.

This is a typical big data application: Doctor McGregor can gather more data than ever before through a more comprehensive sensor. She also accepts that, in this case, not all data are accurate, which can also lead to inaccuracies in her analysis. She put the question "Why" aside, and in a more pragmatic way to help, she looked for "what", which is a better way to foresee infection.

We should remember that big data can save lives.

Correct use of large data can improve medical and educational standards and promote human development

Wen Wei Po: Will the arrival of the big data age lead to a new round of industrial revolution? How can we look at its value objectively?

Meyer-Schoenberg: Big data will dramatically change all aspects of social life, but the question of whether its value equates to the industrial revolution is not yet well said. I guess I can not, because at the beginning of the Industrial revolution in early 19th century, economic development is still at a very low level, so relatively speaking, the people from the process of industrialization of the living standards of the promotion is very huge, today is very different.

What we really want to emphasize is that the big data age will drive us to fundamentally change the way companies operate and the way we live in society. Large data can increase the ability of human beings to make decisions, and this increase will be substantial. With large data, we are not simply improving economic efficiency, but will save human life and prolong our own life. We will also improve education and promote development. In the same way, we must be careful. Big Data also has a "dark side", as we discussed in the book. If applied incorrectly, large data can also be translated into a powerful weapon. Therefore, we must ensure the correct use of large data.

Wen Wei Po: Have you mentioned the "Dark Side" of the big Data age, and will its arrival deepen the digital divide?

Meyer-Schoenberg: Big Data is a powerful tool. So if we use the wrong way, it may deepen the digital divide. But if we use it well, we believe that big data can improve our lives, especially for those who are less fortunate. At this point, you can think of it as fire, electricity, antibiotics, and so on.

Wen Wei Po: In other words, your value perception of large data is based on a longer period of historical development.

Meyer-Schoenberg: If we look at human history in a very broad perspective, I believe that human beings have always wanted to understand the world. At first, many people's "knowledge" is based on superstition and hunches. The development of knowledge is very slow, people need very deep thinking, and then through the practice of testing, to ensure that knowledge is available.

But even so, our knowledge is still not entirely reliable. For example, in 19th century, Pasteur had been studying the rabies vaccine, when a child who had been bitten by a dog and had been infected with rabies, whose parents feared the child would die, begged Pasteur to try his experimental vaccine. Pasteur did, and the child survived. At the subsequent celebrations, Pasteur appeared as a hero, saving the lives of young children. But is that true? Today, through a more in-depth study, we know that only 25% of children bitten by similar dogs are infected with rabies. So 75% of children can survive even if they use ineffective vaccines. The story tells us that we think we live in a very scientific world, but in fact, we have very little data. A new method of treatment requires dozens of or even hundreds of medical experiments to be tested before being proven safe. But it's still too little and people are going to get hurt because we're relying on too little data. In the big data age, we can say goodbye to the lack of data and make decisions that are more rational, based on facts and of course more reliable. This is the hope of the big data age--better decisions will replace the dubious superstitions and unreliable human hunches of our past.

Wen Wei Po: We see that McKinsey released its report in 2011, speculating that if large data were used in U.S. health care, a year could generate a potential value of $300 billion for public administration in Europe that could potentially be worth 250 billion euros a year. A potential consumer's annual surplus of USD 600 billion can be obtained by the service provider using personal location data; By using large data analysis, retailers can increase operating profit by 60% and the cost of manufacturing equipment assembly will be reduced by 50%. The prediction of "data-creating value" has been very encouraging. In your opinion, is big data just a valuable business?

Meyer-Schoenberg: No, big data can do more. Medical care, as we have mentioned before, is just an analysis of some important signs that a doctor can foresee illness a few hours before a premature baby's infection is evident.

Also, through large data analysis, we can find out which part of the school textbook works best for students, and find out the bad parts. So far, we can only follow the human hunch, that is, teachers themselves judge whether students will have questions in understanding a particular course, but in the big data age, we have actual data to refer to, for example, data show that some pages of electronic books have been read many times, because it makes students feel puzzled, so that we can adjust our teaching materials. This will fundamentally change education.

Or take public policy for example: Inrix is a company that provides navigation software for smartphones and provides real-time traffic data. The reason to do this is because each user has become a traffic flow sensor, the location and speed information are sent back to the INRIX company. This will provide a good service to the customers who are travelling around the traffic jams. INRIX Company has a lot of people's activity data, which will also help city planners understand how people commute, where they go to work, then return and build infrastructure such as roads and railways. This is the most effective application. Saving money at the same time, but also conducive to the management of the whole society.

Wen Wei Po: The role of large data in business decision-making, academic research and even national governance is obvious, but will they benefit from ordinary people in everyday life? Why is there a lot of people in the big data age who claim to stay away from overloaded information and data and return to the traditional community life? The individual living space must change from "simple plane" to "multidimensional existence" to have meaning?

Meyer-Schoenberg: The World that man has been through for thousands of years is a world that produces many ideas based on a small amount of data. The map will be redrawn on the basis of this experience when sailors come back from their long voyage. This is clearly not very precise. The cycle of experimentation and error has made human development very slow. However, when we have very little data, this is a natural result. Today we have so much data that no wonder humans are overwhelmed. But big data is now available to help. If humans are not very good at digesting these too much information, large data analysis can help us filter the information and further visualize it so that we can easily use it.

People have not yet generally had the thinking and skills to match the big data age

Wen Wei Po: Some experts believe that the future of large data is the data of the app (accelerated parallel processing) rather than the infrastructure, that is, only data platforms and infrastructure can not create long-term value. What do you think about this?

Meyer-Schoenberg: We believe that the big data age will need at least the ingenuity of people as much as in the past. At the same time, huge resources are the gold mines of the future, and the rewards of those who have these data are unimaginable.

Wen Wei Po: In the Big Data age, data is transparent, how do we strike a balance between protecting privacy, trade secrets and national security? Is your so-called "internet forgetting movement" the best prescription?

Meyer-Schoenberg: The challenge in the big data age is that we find the value behind the data, so keeping it and reusing it over and over again is often a wise choice. At the same time, the current law protecting personal privacy, particularly in the west, targets a world of traditional data, not a large data world. This requires us to make adjustments in the rules for privacy protection. We recommend that you do this by adjusting the relevant protection rules, and as you mentioned, we can choose to forget the data after a certain period of time.

Wen Wei Po: The Big Data Age is an era of massive data to be dealt with and a mass of useless information needs to be deleted. Is that why you stressed in the delete book that we have to make a choice?

Meyer-Schoenberg: Yes. In a way, the big data itself can enhance privacy protection. Because if there are 1 million data points, a single data point is no longer important, which is very different from the traditional data age. Over time, forgetting some of the data will not disrupt the operation and use of the entire large data.

Wen Wei Po: What stage has the big data been developing in the world now? Has the technology to deal with large data been widely available worldwide?

Meyer-Schoenberg: Technology that manages and processes large data already exists and is not very expensive. But there is one thing that is still very scarce, and that is our thinking-to understand the immense value hidden behind the data and the know-how to extract that value. Today, people around the world do not have this kind of thinking and skill universally, but I believe that in the future, this situation will change. We expect that many universities around the world will provide courses for large data analysis to train the skills needed in the big data age.

Wen Wei Po: All previous industrial revolutions, China seems to be both learners and imitators; Unlike the previous rounds of industrial revolutions, China, which has the largest number of data-producing countries, has started technology research and development at the same time as most of the world's most populous countries. Are you optimistic about China's future development in the new era? Does China have the potential for innovation and leadership in the big Data age?

Meyer-Schoenberg: Yes, we are very optimistic about it. China is likely to be a pioneer in the field of big data. In the big data age, China has many advantages: The Chinese are well educated, especially in math and statistics (which is very important). China is a huge pluralistic society that creates plenty of opportunities to create large data resources and build large data applications. Similarly, for large data to flourish, we also need to match the way of thinking, there is the desire to try new things and continuous innovation, with empirical facts as the basis for our decision. So, like many other societies, the big Data age does make a big difference to China.

Wen Wei Po: We know that the Chinese version of the "Big Data Age" is published at the same time as the English version.

Meyer-Schoenberg: In fact, the Chinese version is much earlier than the English version in the United States, but both are based on the same content. This is only because the United States needs more time than China in the production, promotion and marketing of books. At the same time we believe that China is very important and we hope our book can be published here as soon as possible.

(Responsible editor: Lu Guang)