The application of big data in Internet user systems

But for today’s Internet and mobile Internet, the scale and depth of application of big data are no less than traditional telecommunications, civil aviation and other industries. , or even more than a lot. Therefore, the author still wants to write something to briefly talk about the big data application of the Internet, which is a good way to attract more friends. I also hope that more friends can participate in exchanges and discussions.

First of all, the first article would like to talk about the user system of the Internet. Whether the Internet or the mobile Internet, a great characteristic of itself is the Internet, so we can call it the Internet, or the mobile Internet is a subset and extension of the Internet.

In traditional telecommunications, civil aviation, energy and other industries, the customers and main users of enterprises all have identity IDs. For example, the mobile phone card number registered with the ID card in the telecommunications industry, such as the ID card or passport information used by civil aviation users to board a flight, etc. This information can be used as the basic user identity ID, which facilitates enterprises to identify their users and customers, and conduct follow-up verification. Track and analyze user behavior. The great advantage of user information stored by traditional enterprises is its completeness. Many innate and real basic identity information such as name, gender, age and even place of origin can be easily obtained. On the Internet, user access is anonymous. Even if the registration information used by users when accessing the Internet is real-name, it is mainly provided to telecommunications service providers and public security agencies for filing purposes. Ordinary Internet websites are completely transparent in front of users and are "watched". This situation is most typical in the portal website, the main product of web1.0. In the web2.0 era, the Internet has become interactive, and users have gone from simply browsing anonymously to being able to participate in the production and circulation of information through registered identities. At this time, a very important non-determinative condition in the Internet big data applications discussed in this era was born - the user identity system. Why is it called "non-determinative condition"? Because, before this, a large amount of data analysis could be done, but due to the lack of identification of users, the scenarios in which data analysis can be applied and the data obtained are relatively limited, but this does not mean that big data analysis cannot be done. The birth of the user identity system of web2.0 has enabled the Internet to a certain extent to have the same user identity recording system as traditional industries, and data statistics and analysis can be more accurate and in-depth. Among them, early Internet products represented by PC desktop products such as Tencent QQ and Sina UC should have established earlier user identity systems for the Internet. We can also see that these systems were also inherited when their subsequent web products were rolled out. Come over.

So, what information does the Internet user identity system generally have?

When we open any website, we can see that the registration page requires filling in basic information such as username/email, gender, age, etc. Of course, different websites and Internet products have different levels of user profile refinement. Comparing several popular products now, most other Internet products are similar: 1. In Sina Weibo, users can fill in their nickname, avatar, real name, location, gender, birthday, blog address, email, QQ/MSN, Self-introduction, user tags, education information, career information...; 2. Tencent QQ client can fill in avatar, nickname, personalized signature, name, gender, English name, birthday, blood type, zodiac, hometown, location, zip code, phone number , education, occupation, language, mobile phone...

It seems like a lot, so what will the user's information used by the website be used for?

The author Liu Sande here believes that the main points are as follows: 1. Showing oneself; 2. Used as a unique identity ID to distinguish user identities; 3. Related to search and recommendation; 4. The website itself can do user analysis and user behavior tracking. Putting self-presentation in the first place is because this is decided from the perspective of the product meeting the needs of users. The primary task of user profiles is to present themselves as the user's only identifiable identity. Secondly, the author Liu Sande plans to write a special chapter on the related aspects of search and recommendation in the future. You can simply understand it here. The last point, which is what this article focuses on, is using user identities for data analysis. The main dimensions of user analysis involved are user profile and user behavior. Similarly, user behavior is also planned to be written specifically in subsequent chapters. This article focuses on the analysis of user data.

Perhaps according to some articles in the industry and the opinions of old-timers, data must first be large in amount and secondly have high complexity before it can be called big data. However, the author believes that big data does not necessarily have high complexity at the one-dimensional level, and most of it is composed of the simplest data forms. For example, if a website has 10 million registered users, and each user's profile has 6 valid fields, that would be 60 million valid data. And putting these 60 million effective data through one or several layers of simple statistical overlay analysis, cross analysis, etc. is inherently computationally complex. What's more, today's Internet products, especially social products such as FACEBOOK, Tencent QQ, Sina Weibo, etc., often have hundreds of millions of registered users. The user system itself is a very valuable big data.

[page]

What can be obtained by analyzing the user system?

Of course, the information contained in the registration information filled in by the user is the most basic analysis data. Let’s talk with data, as shown below:

The above pictures are from the Internet

The above data are published by third-party organizations, and they are the simplest one-dimensional data. We can see a lot Comparison of user profiles of websites (some of the data sources cited above can also be in the form of online questionnaires, etc.). For an independent website, the analysis of user data is of course limited to the scope of its own website. After entering the Internet web2.0 era, everyone began to pay more attention to users and user experience. Analyzing the characteristics of the website's own users can better distribute the user characteristics of the website, and facilitate more targeted development based on the characteristics of the website's user group. Corresponding product design and development. For example, by understanding the user's consumption level, etc., we can also better provide users with consumption-related displays and services.

So, does the Internet without user identity information no longer have big data? --User identity system without registration.

Some friends may have questions about this topic, and some may be frightened and think that privacy has been leaked. In fact, the application here is also very simple. In Internet products that focus on display, such as traditional web1.0 portals, data analysis and mining can also be done, and there are also relatively mature solutions. Have any friends ever experienced the following scenario: Searching for cars on Baidu, checking car information for a long time, and an hour later, a "car advertisement" appeared on a book reading website. In fact, even if we are not on these websites Registration, Baidu and other search engines themselves can still identify a unique identity information for the user, although this identity information is only temporary and may only be valid for a few days. However, this is still a unique user identity, but the recorded information is limited, but it still provides great help for user behavior analysis. Interested friends can search "google adsense privacy policy" to learn more about it, and I won't go into details here.

The user information system facilitates a series of big data mining

In addition to traditional Internet desktop and web products, the rapid development of mobile Internet and terminal applications in recent years have basically also There is a complete user information system. Apple has created an app store. The number of application downloads so far has exceeded 25 billion, and each download requires the use of a unique user ID. Through analysis, Apple may know what you want better than your parents - this is The scope of user behavior analysis will be discussed specifically later.

In short, the analysis of user identity and data is the most basic analysis in Internet big data analysis. In the era of Internet big data, the user identity system provides the basis for subsequent user behavior analysis and corresponding enterprise product and service design. It provides a cornerstone and lays the foundation for more in-depth data mining.