Big data in China

Chapter 22 Big Data and Life Change (1)
Cloud Sharing and Cloud Consumption

The benefits that the era of big data brings to our lives are of course obvious.Now, each of us holds a mobile phone, and some of us even have several smart phones; we also have computers in front of us, and we can access the Internet at any time; Obtain data to improve the quality of life and enjoy the fun brought by technology.

The data explosion has triggered life changes.Not only is this flooding our world with more information than ever before, but it is growing at a rate that can be dizzying and overwhelming.This change in the amount and speed of information eventually leads to a change in the form of information, from quantitative change to qualitative change.

☆Cloud sharing - unprecedented speed of information diffusion
The first qualitative change related to life is the emergence of cloud sharing.The popular explanation is: cloud sharing relies on users of various Internet terminals as a "database cloud", and uses various resources they like, including resources in video, audio, picture, text and other formats, in a very unified way Share to the database to form a data node that is independent and connected to the outside world.

At this time, if other users need it, they can find these resources conveniently and quickly through search tools on the Internet, and then share them.Need to worry about safety?No, because the shared resources provided by everyone are realized through a certain security protocol, and Internet technology has provided a sufficiently strong and stable protection mechanism.

This concept appeared in 2009, and giants such as Google have invested in the research and development and promotion of cloud sharing.At present, its development in China is also very vigorous.For example, the "Cloud Push" application is very popular now. When people see interesting information on their computers and want to share it with their friends, they only need to right-click to select this function, and they can push it to their friends' mobile phones or mailboxes in real time. The other party doesn't have to do anything to share the information you provide.

This feature of cloud sharing makes people feel free, so it has a very wide audience and has a huge impact on people's lives.It can be said that cloud sharing is the first life change triggered by big data. It truly connects our lives, interests and other people's needs, forming a common world and overthrowing the previous thick closed wall. wall.

☆Cloud consumption—everyone is a consumer information terminal

What is cloud consumption?It is first reflected in the real-time and unimpeded consumption. It refers to the use of cloud computing, e-commerce and other technologies to integrate group buying websites through the cloud and use cloud logistics to extend and develop the group buying end of the supply chain. A centralized group-buying platform based on the e-commerce cloud is established to achieve the construction of a safe and stable supply chain channel, which not only serves consumers, but also realizes the commercial purpose of ultra-large-scale sales.

For example, when each of us has a smart mobile terminal + mobile Internet + SNS account, the actions people share through shopping search and consumption experience make themselves a real "consumer information terminal".At this time, people can not only get other people's consumption suggestions through SNS, but also share their own consumption experience with others.

Typical characteristics of the era of "cloud consumption":

naked age
Like a stripped saint, corporations have no secrets from the market, from consumers, and are all under (often forced) public scrutiny.This includes the production process of the product, the supply process, the cost and even the moral character of the company's executives.

In the age of total nudity, can anything escape public scrutiny?No!The environment of cloud consumption and cloud sharing is so open and powerful, allowing consumers to have a perspective lens through this environment to monitor enterprises in all directions.

storm features
In the era of "cloud consumption", if a company has a crisis, even if the crisis is not caused by it, it will immediately fall into a nightmare environment, triggering a large-scale "condemnation".Like a storm, its consequences are devastating, extending to the media environment and political environment, affecting a large number of consumers, and causing great pressure.The characteristics of the storm also have four characteristics: first, it has a high degree of participation, and everyone in the society participates; second, it spreads widely, and spreads to every corner of the industry, media, government, and public; third, it has great influence and can quickly become a social hot spot , everyone pays attention to; the fourth is strong destructive power, like an irresistible storm, once the enterprise is not well controlled, the consequences will be extremely serious.

There are some typical examples in China, such as the Sanlu milk powder incident, which has the characteristics of such a storm.

Domino effect
The domino effect speaks of the auxiliary effect of cloud sharing on cloud consumption. When a crisis occurs, due to the extremely fast sharing speed, human flesh searches or collective attacks are triggered: everyone pushing the wall is like knocking down the dominoes, causing a certain industry to The comprehensive crisis of the epidemic will not only affect the innocent, but also have a profound impact on the lives of the public.

These three characteristics first brought great challenges to the brand management of enterprises.First, the connotation of the brand is sometimes not very important, or in a certain period of time it will not be valued by people, but will be replaced by word of mouth.When people say you are good, you are good, if they say you are bad, you are powerless to refute.Second, the manipulativeness of brand management is greatly reduced. In shaping the brand image, it can only be treated honestly, and the risk of lying is infinitely expanded, because people can easily discover the truth through the technical means of big data.Thirdly, enterprises should adopt smarter approaches when carrying out brand management. They should neither do nothing, nor be too hypocritical and exaggerated.

There are both opportunities and challenges.But one direction is certain, that is: the core values ​​of the enterprise will become more and more important.It doesn't matter how you say it, what matters is how you do it.

A good company must have its own core values, otherwise it cannot survive in the fierce competition.For example, Apple in the United States is a lofty example that makes almost all companies in the world elusive.

Our "Personal Customization"

When we watch a video website, we usually see a "guess you like" option on the bottom bar or side bar of the page. If you look carefully, you will find: Huh?I still really like it.These websites seem to have the ability to know our preferences, and everyone is "customized". For example, if you like variety shows, there are variety shows in "Guess You Like"; It's a horror movie.

Big data has entered the era of "humanization", and it is also an era of "private customization".Not just "guess you like it", but "I know you like it".

However, with regard to big data, for more ordinary people, they only know this abstract concept. As for what is going on with big data, most people feel more like "heard of it, but don't know much about it".

With the development of big data to the present, ordinary users can also use the entrance of big data.At the end of 2013, the video website iQiyi took the lead in letting users "taste something new". Practical fine print: Green Mirror.After clicking, you can see a collection of edited program clips, which is like a condensed version of a TV series. Green Mirror cuts out the plots you may not like for you.Take the most popular variety show "Where Are We Going, Dad" as an example. The total duration of the first episode of this show is 90 minutes. After you turn on the green mirror, you will see a 29-minute highlight version.Is this elite version edited by the website's video editor?Of course not, it is done by the background system, or it can be said that it is done by all ordinary users who watch the video.Because when we watch a video, we usually click the pause, fast forward or rewind button according to our own interests. Don’t underestimate these actions. The quality of the video content is reflected through these unconscious click actions.

"Use the mouse to reflect what they think is good-looking and what is boring." Tang Xing, chief technology officer of iQiyi, said.

As for the green mirror, it is a summary of all users' "evaluations". The background system undergoes a series of calculations to filter out the dislikes and edit out the most popular ones, thus forming a refined version that meets the tastes of most people.

The "guess you like" we mentioned at the beginning is a traditional algorithm. Green Mirror has made innovations in this area, which is simpler and more practical than the traditional method. The number of user clicks on the first day exceeded 20, and it was in the absence of under the premise of publicity.We can predict that more such big data products will appear soon, which not only saves users' time, but also provides a strong basis for program producers, and they will more accurately grasp people's favorite content.

The era of private customization has come, and it is coming quietly.The birth of Green Mirror actually originated from "Tucao Boss". At first, it was just a joke made by a few engineers. In February 2013, Hunan Satellite TV's program "Everyday Upward" invited Gong Yu, the CEO of iQiyi, as a guest. At the same time, Huo Jianhua and Chen Qiaoen, the two leading actors of the TV series "Swordsman", were also invited.The next day, the show was launched on iQiyi. Out of some curiosity, several engineers secretly used algorithms in the background of the website to generate user behavior data for this episode of the show.

As a result, as everyone guessed, when their boss appeared, the show's ratings curve plummeted to the bottom, and after the two leading actors of "Swordsman" appeared on the stage, the curve suddenly rose to a peak.The ratings curve reflects a very "cruel" problem. Whether a guest is popular or not does not need to be polled, just look at the curve at a glance.

It was this unexpected observation that became the source of inspiration for the development of Green Mirror.It took more than two months for the engineers of iQiqi to finally come up with the algorithm of the green mirror.

Usually, the research and development of a product requires the division of labor and cooperation of several departments.Take "Where Are We Going, Dad" as an example. If you want to produce a green mirror version, you first need to collect data, then clean it, and finally build a model based on the cleaned data to complete the video editing.The cleaning here is like passing through a sieve, which is to filter out those "noise" data, which are not real user behaviors.For example, if a user pauses at a certain point for a few seconds, he may be taking a screenshot, and this kind of data is meaningless; and if he stays at a certain point for a few minutes, then he may be going to the bathroom.

According to the developers of Green Mirror, when modeling, they will have different weights for fast forward, rewind, pause, share, comment, and screenshots, and then score each clip according to the importance. This process is The most complex, and the subsequent calculation process is very simple.For example, a 30-minute video can be finished in only a dozen milliseconds.

Compared with traditional industries, the Internet is much easier to collect user data, because all user interactions will leave traces on the Internet, and these data are not erased, but stored in the huge databases of various companies.Just like Green Mirror, the prerequisite for generating a Green Mirror version is that a program must have more than 10 views, which is easy to achieve.

Usually, the unit of the overall broadcast volume of those first-line variety shows is 20 million, such as "Where Are We Going, Dad", due to its extremely high popularity, it can basically meet the number requirements for generating green mirror versions within an hour of going online.Moreover, every [-] minutes, the background system will be updated again.

Since big data has just emerged, and Green Mirror is still immature, the Green Mirror version of the program has an obvious problem-it is not coherent enough.That is to say, although the video clips are cut out, they cannot be played smoothly.Due to this limitation, Green Mirror can only be used in variety shows for the time being.Variety shows do not have too high requirements for the continuity of the plot. The audience mainly watches their favorite topics or performers, and it will not be affected if some clips are all intercepted.TV series or movies have high requirements for coherence.

Of course, the green mirror version is still being optimized.Since the coherence of the content is not easy to solve, the R&D team focuses on meeting the time needs of each user.If you are a busy person, you may only need to understand the general content through the green mirror version, but for those viewers who have a lot of time and pay special attention to a variety show, you need to watch the full version.

(End of this chapter)