"Big Data": Pursuit, Reflection

You're the target of the big data lock./CFP

-Morning News reporter Miaoshari Lei

"Big data" seems to burst into people's lives. "Probably just like the movie" Matrix, "every element, including every room, every bed, the check-in of each ticket, the location of each car, the swipe of each spot ... A system is enough. "Where to go," said Zhungchengsu, the net chief executive.

Su Ning: "Predictive Marketing"

"Do you need XL or L's diapers? I can sell you a price that is 10 yuan lower than the market price." "This sentence is not from the salesman's mouth, but from the buyer is to buy the delivery of Su Suning easy to purchase the distribution of small Zhang." Despite a little surprise, Miss Xia eventually bought extra diapers. May, Suning selected 5,000 customers in Shanghai and Beijing, based on "Big Data", the first low debug water "predictive marketing."

Xiacian is an easy to buy suning members, the previous 6 months in Suning easy to buy a brand of powdered milk. According to its browsing records, she has been browsing 1 months or so of infant clothing and other products, and the last 3 months, every month in suning easy to buy diapers. On the June 2, Ms. Xia bought three milk powder and other products for 1-and-a-half-year-old children online. This time, the distribution clerk Xiao Zhang specially carried the brand diaper which she often buys, this is the system according to its buying custom specially prepares, this made Miss Xia quite delighted.

"Through the classification of user data and related needs analysis, determine what products to bring to the door." "The analyst may judge the demand product by reference to previous paid orders, product search traces, adding shopping cart merchandise, a single item browsing time, and so on," said Lei, director of Suning.

In fact, through the customer "read Heart" and then forecast marketing, in foreign countries have been tried. According to the test results of the day, more than 260 customers through this "predictive marketing" generated two times consumption, accounting for more than 5%. This ratio, after large data analysis, is much higher than the 0.6% ratio without analysis.

"We want to use the store advantages and large data analysis to make personalized O2O fusion." Su Ning says.

Baidu: Intelligent Mall forecast traffic

5th, Baidu and Vanke reached strategic cooperation, based on large data analysis and cloud computing technology, for Vanke's commercial real estate to provide intelligent upgrade solutions. The first project--Vanke Plaza, Beijing's Changping district, is expected to complete its intelligent upgrade by the end of August, becoming the first Intelligent Mall (shopping center) of Vanke. At that time, customers can search the phone for all kinds of real-time information, such as whether the surrounding traffic congestion, parking lots and how many vacancies. At the same time, Vanke with the help of Baidu data mining, analysis technology, according to customer preferences for the adjustment of shop distribution, investment strategy.

Zhang Dongchen, vice president of Baidu, said the project is expected to be ready for data testing by the end of July and initially on line at the end of August.

Baidu is tireless in the field of large data. This year, from the release of Baidu In-side Innovative Intelligent Hardware cooperation program, to "Baidu Brain" external exposure, and recently please "Google Brain father" Wunda. This April, Baidu opened up large data-engine capabilities, including open clouds, data factories and the Baidu brain. Open cloud solves storage and computing bottlenecks, and developers or entrepreneurs can put data in Baidu's cloud; data factories, which provide "big data excavators". Baidu and CCTV have done "Baidu migration activities", according to 7 billion times a day positioning information, on the map visually show "Spring Festival" during the period of people how to move.

In the hardware field, Baidu and Third-party companies, the search, map and other technologies implanted into the hardware products, to obtain more data. Wang Jin, senior vice president of Baidu, for example, now has a lot of wearable equipment to monitor health conditions 24 hours, record blood pressure, heartbeat, sleep status, movement status, and the detection data uploaded to the large data center.

Ali: "Aliyun" Control the Weather

Alibaba's big Data application is on another level.

Its "Aliyun" and the China Meteorological Bureau Public Meteorological Service Center reached a strategic cooperation, the two sides plan to jointly explore the past 60 years of historical meteorological data: including precipitation, temperature, wind direction, ground icing, solar radiation, acid rain, air visibility and so on.

Where is the value of these "big data"?

Let's look abroad first. Through the analysis of the "big data" of the weather, Merck has more than half a year to master the March meteorological information in the United States, and predicted that warm air will bring allergic factors such as pollen, increase the publicity and supply of allergy drugs, resulting in millions of of dollars in additional sales. Roadside chain stores, depending on the weather changes in the display of display goods, such as in the near rain, the umbrella placed in the eye-catching position.

Analysts said that Taobao every day to produce a large number of business-related data, and weather data matching, the future will be developed for different product sales index, such as down jacket index, sweater Index, mask index and so on. According to weather data, businesses can predict the next six months of hot goods in some regions, advance production and promotion.

[Journalist's notes]

"Big Data" is not all worth it.

Is big Data omnipotent?

Some netizens have complained: a few days ago in Taobao bought a hammer, the results of micro-BO found that full screen are recommended in a variety of hammers. "I can buy a hammer for a few years, why do I have to recommend it?" This is the other side of the so-called "precision push" through "Big Data". In fact, the Giants are rethinking.

Baidu founder Robin Li has poured cold water on "big data" for the current stage: Much of the data is worthless, especially for traditional industries, and the problem of too much data is now to be addressed. The Alibaba team has a dedicated cleaning data team, eliminating some "brush list" error data.

Is there privacy in the big data age?

Think about the day your home address, phone number, body condition, family status, credit card repayment, asset situation ... have become a part of the business, this feeling will not be very good.

Alibaba large data business responsible person said, large data in foreign cooperation, generally are desensitization data, that is, only to tell the results of data analysis operations, not the process. Privacy is above all results. Alibaba has a dedicated data security team, the privacy of the issue, "Data permissions are very strict, other business units can only be based on business needs to apply for appropriate data." For example, sensitive fields, ID number, to check the number of users 1986 years, the data results will not see a person's full ID number. In addition, if the business unit uses the data for different purposes than the original application, we will also be responsible.

"Big Data" is not everything.

Zhi Wei, an industry insider who has done large data models, says big data is valuable but has a time limit. For example, some data is worth looking at right now, but after a year the value is lost.

Baidu has been actively exploring the big data field, CEO Robin Li said that the data produced in the past two years accounted for 90% of the data in human history, "when traditional companies come in the big data age, they think big data is valuable, but a lot of the data we see every day is basically worthless". The next direction, he says, should be "new data," and the challenge is not to create new hardware, but to find what data is valuable to you. Companies can collect "new data" to get truly valuable data: The data can tell you something and may not tell you right away, but sticking to it may be worth it.

Hongyuan Securities [0% Financial Research Report] deputy Director of the Institute Yi Huan said that for China, many of the previous decisions were taken on the head, and now it is important to have data thinking in decision-making. At the same time, in the data accumulation, do not leap forward, but a targeted step by step. such as small businesses, the key is to increase the information construction, pay attention to the collection, use and management of data; In the enterprise, do not throw the data in the warehouse, through the use of data to explore new business model, while large enterprises, the need for lower cost to obtain data, build data assets, to achieve a variety of scenarios.