Liu Xinglong suspected that he might have been "killed" with big data.
This environmental engineer often runs all over the country. In October 2017, when using a certain online car-hailing platform service on a business trip with a few colleagues in Hangzhou, they discovered a strange phenomenon: At that time, they arranged two private cars to depart from the Sheraton Hangzhou Xixi Hotel at the same time. They went to the client's company, because the route was the same and arrived almost at the same time, but when the final payment was made, he paid 35 yuan, while the colleague only paid 25 yuan. The charge for the two special cars is the same.
Why is there such a difference? Liu Xinglong remembered that his account on the online car-hailing platform belongs to a gold card member, while his colleague's account is just an ordinary member.
Afterwards, Liu Xinglong called to complain to the online car-hailing platform, and the customer service staff responded that he could return some coupons, but did not admit that there were big data cracking and targeted prices. Liu Xinglong decided to collect evidence gradually in the future to verify whether the services of various network platforms existed in the practice of "seeing the dishes".
In fact, many people have shared their experience of being "killed" by big data in the online community. On the Internet platforms such as transportation, hotels, movies, and e-commerce, there are many cases in which old users see the prices that are more expensive than new users when purchasing the same network services or goods.
Although many platforms deny that they use big data to "kill familiarity", complaints and revelations of such incidents always occur. What ordinary consumers care about is whether this kind of situation is a common phenomenon in the industry? Is the user's personal privacy data used?
Related platforms "look innocent"
For Liu Xinglong and other netizens suspected of being familiar with big data, relevant online platforms denied it.
"It didn't exist before, and it will never be there." On March 23rd, Zhang Bo, CTO of Didi Travel, the largest online ride-hailing market in China, wrote an article on the Didi intranet denying Didi's existence of "big data." . Zhang Bo said that the Didi platform does not allow price discrimination, and the price will not vary depending on the person, device, or mobile phone system.
As many netizens reported that the estimated price for the same journey is different, Zhang Bo believes that the estimated price varies according to location, road conditions, mileage, and duration. Among them, the road conditions change the fastest. Didi’s estimated price is “refreshed in real time in milliseconds.” The price may vary with the time the phone enters the interface. According to his analysis, some network pictures ignore the coupon deduction, and the complicated network environment may lead to different mobile phone positioning, resulting in different final car prices.
In the actual experience of many users, the suspicion of big data is not only on the transportation platform, but also in the high-frequency field of hotel bookings. Jin Meng (pseudonym) is an assistant to a senior executive of a financial company in Beijing. Because of work needs, she often book high-end hotels on the online travel platform. On March 20th, when she booked a luxury hotel in Shenzhen for her boss, she found that the platform showed that the hotel’s rooms were only 2500 yuan per night premium rooms; but a colleague from the same department told her that on the same platform, she could still see There are ordinary guest rooms for 2100 yuan/day.
Surprised, Jin Meng found that this colleague did not log in to the account like her, but only used the hotel price query function of the platform. Jin Meng is already a senior member of the online travel platform, and since serving as an executive assistant in 2017, he has repeatedly used his account to book luxury hotels in Shanghai, Shenzhen, Paris and other places.
In fact, such things are not uncommon. Prior to this, there have been many "differential pricing" incidents in the United States. When Amazon set a price for a batch of discs, the purchase price set for old customers was a few dollars more expensive than new customers. Later, Amazon responded that this was just a test of random prices and refunded them to high-priced customers. Difference.
Wang Wei, director of the Information Security Department of Beijing Jiaotong University, told the China Youth Daily and China Youth Online reporter that big data is not new. He knew about such incidents two or three years ago. He bluntly said that it is technically easy to master big data without any difficulty. Big data technology can achieve "a thousand people with a thousand faces". Pricing for users of different membership levels was realized around 2013, but now it is not refined enough.
According to him, communication operators may also "openly kill them" in other ways. "For example, as a user of a certain operator, if you don't want to change your phone number, the phone bill will be (a bit more expensive). If you apply for a new card, it will be cheaper immediately. This is equivalent to another form of familiarization." Wang Wei said that people don't have many opinions on this phenomenon, but they often have great opinions on the familiar phenomenon encountered in the process of e-commerce shopping and online hotel/air ticket booking.
Is this kind of behavior illegal? Cao Lei, director of the E-Commerce Research Center, believes that at present, it seems that this is just a "side-by-side ball" and it is difficult to define. In theory, both merchants and e-commerce platforms have the right to set different prices for the same item, rather than all having to set a unified price. But the familiarization of big data does expose the asymmetry and opacity in the development of the big data industry. The phenomenon and behavior of this case should be severely cracked down.
Wang Wei pointed out that it is inappropriate to use big data technology. Because this behavior not only makes users lose their trust in the company, but also leads to increased user costs.
Is everything transparent in the era of big data?
Jinmeng is very curious how the "thousand-thousand-thousand-face" big data technology can display different prices to different users. At the same time, she is also very worried. Does this mean that everything she does on the Internet is transparent?
Wang Wei analyzed that the prerequisite for big data to be familiar is that the platform must master data such as personal information and behavior habits. For example, a user often logs on to the website to book air tickets and inquire about the flight from Beijing to Guangzhou. When the app is opened next time, the default flight information is Beijing to Guangzhou. This shows that the software has learned user behavior and frequency. In addition to the information that users actively enter, the information that users actively disclose may also be helpful, such as ID cards, phone numbers, and address information that need to be protected by the legal system.
Wang Wei said that they are also conducting research on this information, such as extracting dimensions from Weibo information. Even if the user does not disclose his occupation and age on Weibo, he can roughly figure out his age, occupation, hobbies, and personality based on the photos he posted, where he traveled, and what he did.
"Everyone has about 100-1000 keywords, (these keywords) can portray you." Wang Wei said, for example, if users often fly to Guangzhou, then the Weibo ads that pop up will be fun in Guangzhou. , Where is cheap, etc., the thing pushed is just "put it in your heart."
In fact, the behavior of establishing user portraits based on the user’s personal data, traffic trajectory, purchasing habits and other behavioral information, and then using this to achieve corresponding product recommendations, has long existed. Many Internet companies have dedicated jobs in charge of providing user data. Labels, as far as possible to achieve a comprehensive understanding.
Previously, a research and development staff of an online travel website once wrote that when the website made user portraits, the data involved included payment ability, travel preferences, destination preferences, family composition, website or App page stay time, etc. , The coverage is very wide.
How to use privacy-related data is the key
In fact, how to understand the needs of users better and not overuse data involving personal privacy has always been a problem faced by Internet companies, especially those that occupy a leading position in the market in some areas.
Deng Zhisong, senior partner of Beijing Dacheng Law Firm, said that at present, companies usually have two ways to master user data. One is that the company collects user information when users use company-related products and services. Taking Ctrip as an example, its "Privacy Policy" stipulates that it "may" know the user's "travel plans, styles and preferences" and other information. The second is that companies share user information through database "collision", that is, different websites exchange databases, or share user information through certain transactions.
According to the "Consumer Rights Protection Law" revised in 2014 and the "Cyber Security Law" implemented in 2017, the consumer's consent must be obtained to collect consumer personal information. Deng Zhisong believes that many businesses now use databases to "collide" to obtain more user information in order to make user portraits more accurate. However, if the collision does not obtain the prior consent of consumers, there will be an illegal risk of collecting personal information without user consent. .
Deng Zhisong bluntly said that big data is a kind of price discrimination. Faced with consumers with the same trading conditions, companies use low prices to attract consumers who use websites with low frequency, while charging high prices to high-frequency consumers. If the company’s market share exceeds 50%, differential pricing may be suspected of constituting a monopolistic behavior that abuses its dominant market position. Deng Zhisong suggested that the protection of personal information should be strengthened in accordance with the "Cyber Security Law" and the "Consumer Rights Protection Law", and verify that businesses have clear consent and authorization as a basis for obtaining so-called "big data" related to user transaction records and habits . If the market share of businesses that have "killed familiarity" is relatively high, the anti-monopoly law enforcement agencies should also intervene in the investigation.
In Cao Lei's view, big data is a "double-edged sword". The evasion of the phenomenon of "killing familiarity" still depends on the self-discipline of the enterprise and the government's control. "Pricing pursues fairness and justice, and we must publicly declare special prices." . What does the regulatory authority do? Cao Lei believes that the act of killing familiarity with big data involves a wide range of areas and requires many government departments to participate in supervision, such as industry and commerce, commerce, transportation management, industry and information, and Internet Information Office, etc., involving multiple regulatory agencies, and the powers and responsibilities are very unclear. . At present, it seems that the first thing to do is to clarify who should be responsible for all kinds of problems and all links.
