How is big data shaping us?

The value of big data learning is to mine and analyze users' behavior habits and preferences from behind the complex data, to find out products and services that are more in line with the user's "taste", and to adjust and optimize itself in a targeted manner based on user needs .



The concept of metadata (Metadata)



Simply put, metadata is the data that describes the data itself, it is not the object itself, it only describes the properties of the object.



For example, a painting itself is data. And the painting's author, completion time, size, price, type, etc., is its metadata. The value of metadata, the first is that it can describe the object from the side; the second is that it can be structured and informationized.



What does it mean?

For example, if we want to judge the value of a painting, in addition to experts directly evaluating the artistry of the painting, we can also judge it through metadata. Is this painting by a famous artist or a second-rate painter? Is the painting the work of the author in his heyday, or when he was young? Is the painting a genre the author is good at or is he unfamiliar with? Use these descriptive information , we can count the value of this painting as close to ten. Although there will certainly be errors, it is also a scientifically sound method.



So what's the point of describing objects with metadata rather than the data itself?

This is the value of big data: for unstructured and unquantified objects themselves, structured metadata can be used to quickly calculate and judge.



The first stage of big data application: auxiliary products The



initial application is relatively simple, which is to assist product personnel and market personnel to make judgments.



In the past, it was very troublesome to do a research on physical products. For example, in a beverage company, researchers use a variety of ways to watch the scenes and steps of their drinking. Questionnaires are the most common, but not accurate enough. Therefore, various professional field tests will be organized, an environment (usually with single-sided glass or a camera) will be set up, volunteers will be invited, and then they will be guided to complete some operations according to their daily habits.



Obviously this approach is very cumbersome.



Today's Internet products don't have to be so troublesome at all. All usage data and behaviors of users are recorded, and what they want to know can be analyzed in an instant. In the past, it was very difficult to know whether users have done one thing and whether they have used this function.



Now, I want to know how many times the click was clicked, where it was clicked, when it was clicked, and even where it was clicked. Usually users use this function or not, how to use this function, it is clear at a glance.



For product designers, this is vital data. Moreover, this is complete data! If it is an Internet product, then what I know is the data of all users, not the sample data of traditional industry products in the past.



Tencent knows how many WeChat users use Moments, how many Moments these users post every day, and what these users post every day. Every data is real and available.



(In the past, it was difficult for newspapers with large circulation to know the gender of readers, but now no matter how small the WeChat public account can be obtained in real time)



In the industry of physical products, with the informatization of production, sales and use of the entire product in the future, large Data will gradually play a bigger role. A bottle of water I sold in the past may have been broken at a supermarket. I don't know who bought this bottle of water.



But now that I sell a bottle of water on Tmall, I know that the other user buys ten boxes of water every month, and his address is a high-end restaurant, so I know who the target audience of this bottle of water is.



This is the value of metadata.



Therefore, the first stage of big data is to assist product designers in making judgments and allow product manufacturers to better satisfy users.



At this time, big data is mainly used to provide support for products, which are then applied to users.



The second stage of big data application: creating value



After the quantity and quality of the data reaches a certain level, things start to change. Metadata will not only serve as an aid to the product, but become the most valuable product itself.



Very simply, is the Bureau of Industry and Commerce most familiar with the consumption habits of ordinary people in China? Which association? Which scientific research institution? No, it is Taobao.



Is it the Personnel Bureau that has the most comprehensive personal credit information? Is it a bank? Is it a consulting company? No, it is Alipay.



The reason is actually very simple. All behaviors (consumption, transactions) occur on this platform, and this platform has records of all data, then these data can generate huge value.



Do you think that products in the field of medical and health care are only concerned with your health? No, they can also record all your physical signs, which is the first-line clinical data.



At this point, big data itself has become a product that can output valuable content.



Consumer behavior data can be sold to advertisers, and advertisers can target you with advertisements; credit data can be sold to banks, and banks can determine your credit level. Health data, sold to insurance companies...you get the idea.



Not just selling data, but what data provides can create new products. Especially for products/services such as O2O, the upstream is the service provider and resources, and the downstream is the user, all of which can be valuable and can be discovered.



When doing manicures in the past, the business model we envisaged, the upstream is to understand the situation of manicurists' supplies, cooperate with manufacturers, and control the channels; the downstream is to understand the situation of users, so as to be able to cooperate with other beauty products (directed to help you). Bring the product to the home, which Beaver is already doing), to make user data generate value.



I heard before that Ele.me is experimenting with a new service, which is providing ingredients to restaurants. It sounds a little weird at first, but then it makes perfect sense. Besides Ele.me, who can better understand the food sales data in a certain area? How much and how much radish and cabbage are sold in this place? upstream channels.



Big data at this stage can already become a product, serving users directly.



From another perspective, through the metadata of our behavioral data, we are slowly being described by quantitative information. Seeing these numbers (how much money is spent in a year, where it is spent, etc.) can already give a relatively rough idea of ​​the person.



The final form of big data is beginning to emerge.



Stage Three of Big Data Applications: Shaping Us



I've always been disdainful of behavioral data. Do you know who I am when I buy something on Taobao, chat with someone on WeChat, and check something on Baidu? It



’s really possible. As long as the data is of high quality and quantity.



I found that you chatted a lot with a girl in a different place on WeChat and often returned videos, so this is probably your girlfriend in a different place; I understand that you have been searching for air tickets and travel strategies in Southeast Asia on Baidu, then I know you may be going. play there.



With just a few simple pieces of metadata, I can infer that there is a high probability that you will travel to Southeast Asia with your girlfriend recently.



This is to illustrate the logic of metadata being able to reason about information. And there are more and more metadata available.



As the owner of the data for these products, I have absolutely no need to send a private investigator to follow you. Just wait for you to submit the data yourself.



During the Spring Festival, why did Alipay compete with WeChat for micropayment and social payment? Not for the handling fee, but for the social payment that it lacks. The value of this piece of data is far beyond imagination.



In the future, each of us will have a large amount of data records on our daily life. Our actions turn into strings of numbers that become quantifiable data, the information that describes us. We use cloud notes for work, Ele.me for meals, Didi for taxis, Baidu for search, and WeChat for social networking. Every step is recorded in detail.



If you don't believe me, you can dig out all the search records in your history on Baidu or Google. The description of your life is definitely more real than your own diary.



This data will be converted into valuable business data to describe all aspects of your information.



Ultimately, we are ourselves big data objects



that are pros and cons to a future like this. The advantage is that we enjoy the convenience brought by big data everywhere, every advertisement we see will be our favorite, every search record we check is recommended according to our characteristics, when we add friends The system can even say if he will get along with us.



The downside is that our privacy is exposed. As long as the owner of the data wants to do something bad, anything is possible.



Big data will never stop at providing help for decision-making. Brothers (www.lampbrother.net

) big data training, its ultimate form is that it can describe our specific individuals with massive data. When this step is reached, the current market research and user analysis will become extremely simple.



Because big data has completely shaped us.