Zuckerberg's law--talking about the big data age

Less than a year ago, Facebook founder Mark Zuckerberg announced at the Web2.0 summit that social-sharing information was growing exponentially, according to Facebook statistics, and that today the total amount of sharing was twice times higher than it was two years ago, and that the total amount of information sharing that users would generate would double from the year that began. Zuckerberg's social-sharing law can be expressed as a formula: y = cx2x. where x represents time, Y represents the user's information-sharing, and C represents the current moment of sharing the amount. If this formula is established, then 20 years later, a user's information sharing will be more than 1 million times times today, that is, 2 of 20 times.

The introduction of this law is a summary of the changes in the amount of data shared by the network user behavior in the WEB2.0 era, including, of course, the business purpose and the imitation factors of the predecessor of the information industry. If this law is true, Facebook's prospects are limitless. Even according to the company's current, simple, traffic-based advertising business model, Facebook's annual income can be endless. If the existing business model is optimized and new revenue methods are introduced, the company's annual revenue growth will continue to grow at a high rate for a long time. Xiao Zha is a very clever way to use this straightforward law to do PR and to trick investors and partners. This is obviously inspired by the Metcalfe Law of information industry for many years, this law says that the value of the network equals the square of network nodes, the value of the network is proportional to the square of the number of users connected, that is, the value of the network v=kxn2, where K is the value factor, n is the number of

Is Zuckerberg's law really tenable? If so, how do you understand it? This law cannot be established if it is viewed from a simple and intuitive experience for individual users. Only 24 hours a day, the time for everyone to surf the internet is limited. How can a person face such a large amount of data in a finite amount of time, if 10 years later, his network behavior contributes more than 1000 times (2 of 10 times) and 20 years later? Can't imagine a person reading ten years after the volume of news is thousands of times today, or 20 years later, a person to visit the online mall or play online games time is today's 1 million times-fold. Similarly, if the company that runs the traditional Web site, the amount of data they can generate and use can only be increased by the number of levels, and Zuckerberg's law is not a point, it is impossible to cheat investors.

Zuckerberg's law can only be set up on sites such as Facebook, which is only suitable for Web services in a WEB2.0 architecture that is user-centric and interacts and spreads through human relationships. For example: According to sociological research, people can effectively maintain 145 social relationships on average, meaning that 900 million of Facebook users can have as many as 145 friends per person. If a user and his friends only say a word or have a move today, he will get 145 shared information from a friend, and the message he produced was spread to 145 people. If the user and his friends say 10 words or 10 actions tomorrow, he will get 1450 messages from a friend, and the amount of shared information generated in the Web site is 145 10 times. In other words, if the network behavior of each user is exponentially growing, then the communication, interaction, and sharing of information between users is a geometric growth. Compared with the traditional web site, for example, on a portal site, a user read a news only to get a record of behavior, more than 10 news to get more than 10 records, the amount of data does not produce geometric growth effect, Zuckerberg's law failed.

Facebook's empirical data is that it doubles the number of shared data each year, but theoretically the growth of shared data should be doubled. So, this law is not an exact mathematical law, but a statistical description of a trend. The growth of shared data may vary from one site to another in different historical periods and at different operational levels. But the big trend is not wrong, this law will be valid for a long time. The famous Moore's Law of history was proposed in 1967, but still valid 45 years later today.

Recall that in the 80 's when the study of social networks in the United States, the greatest pain was not in the grasp of the theory, nor in the study of statistical models or computer programming, but in the nowhere to obtain large-scale, systematic, complete, dynamic social network data. But only by using computer simulation or some local simple data as the basis for analysis and research, so over the years it is only a small circle of things to entertain themselves, the development of unpleasant, little impact. Until the last few years, exactly until Facebook was born, social networking analysis is, to a large extent, the test of thinking, not empirical, with economic or physical theory. Just as the advent of computer science in the Second World War has facilitated the independent emergence and flourishing of the university's computer sciences, the development of services based on social networking mechanisms on the Internet may promote social network analysis as part of sociology. Maybe in the near future, sociology will break free from social philosophy and empirical social research, and become a science subject like economics, Jeannobert Award adds a social Science award.