How does big data know that you have been to Xinfa?

The Xinfadi wholesale market

“After the city’s big data analysis, you may have visited the Xinfadi wholesale market after May 30 (inclusive)…” As the source of the new crown in the Xinfadi market is locked, in recent days, big data screening has become a lot A topic hotly discussed by Beijing citizens in the circle of friends.

A set of data shows that as of June 17, under the condition of consistent action trajectory, using big data to analyze the position and path of the population at risk, Beijing spent just a few days to “find out” more than 300,000 people. Nucleic acid amplification testing. Compared with the population movement of big data analysis in the early stage of the outbreak in Wuhan at the beginning of this year, the “big data screening” in Beijing this time undoubtedly embodies a higher technical content and better timeliness, which has played an important role in the treatment of the epidemic. effect.

The picture shows the people who visited Xinfadi from May 29 to June 12, and found that 2430 people visited the Xinfadi market and left Beijing, and 91 of them visited Shanghai. The chart also lists the other cities that more than 2,000 people have visited, including Langfang, Baoding, Tianjin, etc.

Precise search

Three ways to target new contacts

Since June 15th, many Beijing residents who have passed or visited Xinfadi have received short interest rates or community calls, indicating that through big data screening, they have confirmed that the respondents have recently visited or passed the Xinfadi market. , Request truthfully fill in the information, and perform nucleic acid detection as soon as possible.

So, what is “big data screening”? How does big data find new passersby? What role did big data play in the prevention and control of the epidemic?

Big data, as the name implies, is actually massive and large amounts of data. These data are derived from data generated at any time; and big data is divided into narrow and broad categories; big data in narrow sense includes personal portraits, shopping habits, reading habits and other personal portraits ; Big data in a broad sense is aimed at society or enterprises, for example, e-commerce uses big data to analyze customer shopping habits, make demand forecasts, and arrange warehouse storage in advance.

Obviously, the newly launched “big data screening” is an organic fusion of generalized big data and narrow sense big data, and is also a good use of mass storage and fast retrieval technology.

Although the official did not announce the specific method of big data screening, Dr. Song Huaiming, chief scientist of the chief engineer of Zhongke Shuguang Big Data and deputy general manager of the storage product division, analyzed that Beijing has been able to use big data for rapid screening and screening of new places Visitors, there are three technical paths that can be achieved.

The first is to determine the location information based on mobile data of the mobile phone, also known as the base source positioning method. This is the most commonly used method and the most efficient method in this new big data screening: but the base source is not GPS, but Through the mobile phone base station, it is different from the GPS when it is turned on. The mobile phone will automatically connect to the nearest signal transmission tower. All activities of the mobile phone can be traced back through the signaling data signal transmission tower containing the base station information. This is used for tracking. The location and tracking of the person’s location provide true and accurate first-hand data.

Secondly, through social communication information analysis, this does not mean judging through social software information, but combining existing data, through telephone surveys, random visits, etc., to form relatively reliable data information, which can also be screened out in a short time Who works in Xinfadi, who has been to Xinfadi, who they have been in close contact with in recent days, etc.

The third method can be confirmed by the item information. In this new site screening process, in addition to the screening of passers-by and close contacts, many items and goods are also checked. In this process, through the specific The path of carrying viral items and contact tracing can also be used as a basis for checking and finding close contacts in new places. However, Song Huaiming emphasized that the data of this screening method may not be complete, and whether Xinfadi really uses the method of searching objects to find contacts is currently not entirely certain.

In addition to the official official response, there was also a saying on the Internet that “Alipay and WeChat provided data, locked 350,000 people, and helped virus screening”. But the news was soon officially rumored by both parties.

In this regard, Song Huaiming believes that the transaction user’s location can be determined by acquiring transaction records through a QR code. But for the tracking of the transaction path of transaction users, neither Alipay nor WeChat can be achieved. In addition, as a bulk wholesale market for agricultural products, Xinfadi has many other transaction methods through online transfers or cash. Therefore, Alipay WeChat data can only play a certain auxiliary role in big data screening.

Technological innovation

Big data helps precise prevention and control of epidemics

In addition to screening visitors, what role does big data play in this new epidemic? Wu Zunyou, chief expert of the Chinese Center for Disease Control and Prevention, told the Beijing Science and Technology News that in his view, in the process of the new crown epidemic in Beijing, the big data played at least two roles: first, accurately lock early cases to the new place In the market, discovering the source of infection in time has won valuable time for disease control.

The second is the understanding of the outflow of people who have traveled to high-risk areas after the outbreak was discovered. This includes the exposure to the environment, the contacts and close contacts, the distribution in Beijing, and the situation of leaving Beijing. Potentially infected people play a very important role in blocking the spread.”

In fact, not only in the new place of Beijing, in the process of this new coronary pneumonia epidemic, especially at the beginning of the outbreak in Wuhan early this year, big data screening and analysis have played a very important role in disease prevention and control.

Beijing has achieved accurate grading this time, there is no'one size fits all' to seal the city', but the risk level is divided by street as a unit to carry out grading management and control, and big data plays a key role

After the outbreak, several technology internet companies have provided a large amount of data support for the whole society through data and technical capabilities; for example, Baidu’s migration data accurately calculates the number of people leaving Wuhan and staying in Wuhan, providing a reference for government prevention and control. Another example is the 12306 ticketing platform, which takes advantage of the big data of real-name ticket sales and cooperates with local governments and prevention and control agencies at all levels in a timely manner to provide close contact information for confirmed patients on the vehicle.

In addition, using big data technology to achieve information sharing and fast query, many media platforms and Internet platforms have opened up new crown pneumonia dynamic updates, epidemic situation display and rumor dissemination functions. Users return to rationality among the mixed information of the dragon and the dragon, allowing the authoritative information to “run” in In front of the rumors, the public sentiment was calmed in time, and the spread of the virus was “suppressed” from another level.

At the prevention level, early warning can be given to the gathering of people; in terms of the source of diffusion, combining mathematical models and social characteristics of personnel can analyze the diffusion path and speed, and formulate more targeted countermeasures and measures. “Beijing has achieved precision this time. There is no “one size fits all” to seal the city, but the risk level is divided by the street as a unit to carry out classification control.” Song Huaiming believes that this is where big data plays an important role.

Compared with domestic, in terms of big data prediction and analysis of epidemics, foreign countries have had more mature exploration and practice. As early as 2008, Google released “Google Flu Trends” (Google Flu Trends), using keyword tracking technology Collecting data, if there are a lot of searches for keywords such as flu, fever, and cold within a certain period of time in a certain area, it indicates that there are potentially infected people in this area, which needs to be paid attention to by relevant departments.

Google once developed software that uses big data to analyze flu trends, but it ended in failure.

In 2009, when the swine flu broke out in Mexico, researchers also used communication data to monitor the public’s response to the government’s health warning information to guide relevant authorities to better adjust policies; the 2014 Ebola outbreak in West Africa, Researchers also modeled and analyzed the prevalence of infectious diseases based on mobile phone communication data modeling.

It is not difficult to find that with the public health field supported by big data + artificial intelligence: on the one hand, it is more efficient and faster in action, effectively slowing down the spread and spread of infectious diseases; on the other hand, it can also play a role in providing prediction and The important role of analysis.

Plan ahead

How big data applications are “higher, faster, stronger”

After the outbreak of the new crown epidemic, big data was quickly applied in epidemic investigation and other aspects. Domestic fields, industries, institutions, and departments have used big data technology to strengthen interconnection, accelerate the integration of the entire society, and play an important role in the treatment of the epidemic. The role of gaze.

However, the application of big data is not a perfect body. With the deepening of the epidemic response, it has also exposed many shortcomings and areas that still need to be improved.

The first is accuracy. From the current technical level, the application of big data in most cities focuses on urban daily life management scenarios and basic security scenarios, such as in the financial field, security field, transportation planning field, etc.; but During the epidemic, solutions to “emergency public health incident scenarios” need to be found. There are many uncontrollable factors. It is easy for some conventional algorithms to expose shortcomings and even inaccurate data.

For example, in the epidemic prevention process of this new location, the big data information cannot accurately distinguish the difference between the “visited” and “passed by” new location markets, and due to the short operational time, the “one size fits all” situation eventually appeared. Citizens who took the public transit route to new places also received text messages or notifications requesting nucleic acid testing, which was quite surprising.

In this regard, Wu Zunyou believes that during the big data screening process, due to the short time and limited energy, there are indeed some wrong judgments, and some citizens who are not exposed to danger are also required to fill in information to detect nucleic acids. “But the number is only a small part. At present, the primary task is to control the epidemic, and there is room for improvement in the future.”

Song Huaiming also believes that in addition to sending text messages, screening methods also include phone calls, door-to-door visits and self-reports, etc. The purpose is to minimize the inaccuracies and inaccuracies of big data applications in public health emergencies Determining factors.

The second is the big data decision-making problem. Song Huaiming said that from the actual effect, the current role of big data is still more in analysis and auxiliary decision-making, and only a small part can make automatic decisions. “From a data perspective, this The approach is more scientific, but from the perspective of time and efficiency, such a method still has much room for improvement.”

In terms of data legislation and data security, the relevant laws and regulations are not sound enough: the reporter learned that there are currently no laws and regulations for public data management at the national level, and the provinces and cities are basically in a situation of fighting each other, and some places have introduced laws Regulations, some places have only promulgated rules and regulations, and some places have not made legislation.

This has brought uncertainties to data security and personal privacy in an invisible way. During the epidemic, there have been many incidents of personal privacy leakage, which has caused public concern. For the prevention and control needs, many people’s personal information has been “Exposed” in APPs, applets, registration forms, or notebooks, which are reasonable collections and which are excessive collections, what if personal information is leaked? It is still open to discussion,

In fact, strengthening information sharing, eliminating information silos, and protecting personal privacy and data security are not only important for fighting the epidemic, but also have a huge impact on the realization of scientific government decision-making, precise social governance, and efficient public services. Promote the role.

“In the future, we must focus on regulating the ownership, use rights and usage standards of data, and protect the privacy and data security of the public.” Song Huaiming emphasized that the relevant departments should not only grasp the right of the public to know, and the relationship between protecting public privacy “also As far as possible, release the data that the public cares about in a comprehensive, timely and accurate manner, and fully mobilize social forces to actively participate to form a joint force for big data sharing.”