Large data records is everything an angel or a demon?

Many people feel inkling when they raise big data, but the big data is what it is and few people can explain it. During the two sessions this year, including Millet Technology founder Lei, Lenovo Group chairman Yang, Hkust flying (002230, shares bar) chairman of the Liu Qingfeng, including several representatives, Members have put forward with large data-related proposals and proposals, on the one hand, they hope to promote the development of large data from the national level, On the other hand, it also gives a warning to the potential hidden dangers in information security.

Big Data records everything.

A survey showed that in 2013 China generated more than 0.8ZB of data, equivalent to the total number of global data in 2009, and by 2020, an average Chinese household will produce a yearly amount of data equivalent to half the National Library's information reserves.

Xu Xiaolan, Secretary general of China Electronic Society, pointed out that the huge amount of information storage and mining, is not only the value of large data, but also it is different from the traditional internet, may bring new hidden dangers to information security.

"In the big data age, there are a lot of data in the past that are impossible or not needed to be recorded, such as microblogs, the content of friends, the cookies generated on the Internet, the use of household water appliances, the data from cars and sensors installed on large equipment." "If all this information is an island, the impact may not be significant," said Zhang Wenhao, a unique technology CEO at Social Data Analytics. But once interconnected, the impact is staggering. ”

Tian Jing, director of the Institute of Information Engineering at the Chinese Academy of Sciences, also said that the previously fragmented data were only elephant, but now these fragments are all stored and pieced together through correlation analysis, "We know what it looks like."

The technical gap has also created a temporary "insecurity" of large data.

"Without our own analytical skills, how can we figure out what data needs to be protected?" ”

For this "problem" of Tian Jing, National informatization Advisory Committee deputy Director, Chinese Academy of Engineering Hequan A set of data provided the answer-about half of the data should be protected, but we now really protect the "half of the half", a lot of data are intentionally or unintentionally to be obtained, " If these data are integrated and exploited by some people with ulterior motives, they can have a significant impact on security.

Xu Xiaolan Introduction, at present, many institutions and enterprises in China use large data analysis software are produced by foreign manufacturers, especially in recent years, such as video and other multimedia data explosion, "These data are isomerization, this part of the software is our short board."

Yin, a military expert, pointed out that the Internet is now used by most of the servers in the United States, "I do not have a server how can it be safe?"

On the other hand, in Tian Jing's view, the lack of a large data environment in the security concept is also one of the important reasons. "The information that used to be harmless is completely different today in the data explosion," he said. The idea does not change, there is no secret to protect. ”

Zhang Wenhao also believes that people lack this awareness. "In the United States, a lot of people will ask not to disclose their privacy, or take the initiative to provide their own data to some specific agencies to use, but in the country almost did not hear, you do not know what type of data can be a lot of risk." ”

Big data may also be "dangerous."

Zhang Wenhao that many people's understanding of large data only notice the volume and statistics of the data, "in fact, through in-depth analysis, from the numerous data to abstract the rules and principles, and to achieve a forward-looking forecast for the future, is the real value and charm of large data."

Because of the big data, Obama became the first U.S. president to win re-election in the past 70 years in 2012, with a high unemployment rate of 7.4%, and target supermarkets have "predicted" the pregnancy of a 18-year-old girl.

Behind any new technology, a sword of Damocles may hang. The same is true of large data, and the threat is creeping in when people marvel at its "divine power".

In the previous two years, a foreign intelligence collection organization used the data and information published by a domestic agency to carry out in-depth mining analysis and generate valuable information. Xu Xiaolan told reporters that later after a detailed investigation, confirmed that the information is not actually provided by the personnel, and the information generated by the data itself is also open. "It was almost impossible in the past and it sounded a wake-up call," he said. ”

Yin with wartime traffic flow information, for example, if you do not pay attention to the safe use of data, it will be possible to the national security risks. "During the Kosovo War, the FRY's oil storehouse was well concealed, but the U.S. military, by analyzing the traffic flow in the satellite map, delineated a large number of areas where the tanker often haunts, and then blew up the exact search." ”

If you think these "risk factors" are far away from you, it is a big mistake.

"Many smartphone apps now require access to the Address book," says Hequan, a lot of people who feel they have no secrets and agree. "But in practice, it will not only reveal a lot of information about yourself, but also put a lot of people at risk of privacy exposure." ”

Zhang Wenhao also pointed out that many young people now use mobile phone to send micro-blog or "check-in" habits, "this will actually reveal a lot of their information." For example, you often check in the daytime location is likely to be your unit, the evening after the signing of most of their own home.

Big data is experiencing growing pains

"To conquer the data, only make better use of them." Zhang Wenhao said that data is a resource and wealth, the more accumulation, the greater the effectiveness.

This view is Hequan. He believes that the more you want to avoid the security risks of large data, the more you need to collect data and dig deeper. "Statistics show that domestic data collection is less than Japan's 60% and North America's 7%, a lot of data left in vain." ”

"The data that companies have is limited and local. "Robin Li, chairman of Baidu Company, has proposed the concept of data openness, and the state should disclose data that does not involve safety and allow better analytical use by competent institutions."

The original chief engineer of National Basic Geographic Information Center Li Li pointed out that some public departments have a large amount of public information, which is an important part of national basic information. Hequan stressed that the data "do not want to share with other departments, resulting in incomplete information or repeated investment."

Is there a lot of technical difficulty in sharing data? In Xu Xiaolan's view, there is a need to vigorously develop data mining, analysis technology, but for data sharing itself, technology is not the biggest obstacle, the key or benefit coordination. She said a few years ago, the Ministry of Land and the CBRC prepared a mapping of national land information, "at the outset many people put forward various difficulties, almost considered impossible to complete the task." But later, the introduction of administrative means such as accountability, to get local co-ordination, carried out very smoothly.

Zhang Wenhao that the establishment of an open, transparent, standardized data market, will greatly enhance the utilization of data. But in this process, it is necessary to carefully consider what data can enter the market.

This involves the issue of legislation. Defining "privacy" and classifying data safely is a priority in enacting laws and regulations. "There is a game between personalized service and privacy," he says. Name, telephone number, address ... You can't say anything private, so you need to rank data security. Simply put, through the experience of data analysis, we will know what kind of information is more directivity and exclusive. The level of security for such information should be higher. ”